CROSS-REFERENCE TO RELATED APPLICATION [001] This application is based on and claims benefit of priority of U.S. Provisional Patent Application No. 62/904,016, filed September 23, 2019, the contents of which is incorporated herein by reference in its entirety.

TECHNICAL FIELD

[002] The current disclosure relates generally to systems, method, and devices for farming. More specifically, the current disclosure relates to systems, methods, and devices for autonomous farming.

BACKGROUND

[003] Traditional farming must change dramatically to meet demand. It is estimated that under traditional farming methods, fresh vegetable shortages may occur as early as 2023. To satisfy demand, it is estimated that, by 2050, farming yields must increase by 50%. Global warming is also impacting agriculture resulting in unstable growing conditions, lack of arable land, and water and resource scarcity. These challenges are also resulting in a lack of skilled farmers.

[004] Autonomous and semi-autonomous, indoor farms that can grow eight times the crops using no chemicals, herbicides or pesticides, and with significantly less water than traditional farms have been in development now for some years now, and their popularity is growing. However, these currently existing firms are not economically viable and may be unprofitable. Human labor accounts for 50% of a farm’s expenses. Therefore, what is needed is a solution that cuts those expenses while guaranteeing exceptional produce, no matter the geographical area. Health-conscious consumers have increased demand for fresh, sustainably-grown, locally-sourced strains of lettuce and herbs. Therefore, there is a need for a system of indoor autonomous growing that lowers the barrier to entry for users, and allows for a high level of consistency through a variety of locations, ranging from dense or remote population centers with commercial food and medical needs to consumer households. The current disclosure teaches systems, methods, and devices that rectify at least some of the above described deficiencies. SUMMARY OF A FEW ASPECTS OF THE DISCLOSURE [005] A growing system for growing different types of crops include one or more sets of vertically stacked growing pods arranged on a surface. The growing pods of each set may be configured to grow crops and may be electrically or fluidically connected in series. Some embodiments of the growing system may also include an autonomous rover configured to transport a set of stacked growing pods from one location to another and/or a robotic forklift configured to unstack the growing pods of the first set.

[006] The foregoing summarizes just a few aspects of the disclosed embodiments to provide a brief flavor of the disclosure and is not intended to be restrictive of the inventions described, illustrated, and claimed hereafter.

BRIEF DESCRIPTION OF THE DRAWINGS [007] The accompanying drawings, which are incorporated herein and form a part of the specification, illustrate embodiments of the current disclosure. These drawings together with the description, enable a person skilled in the art to make and use embodiments of the current disclosure.

[008] Fig. 1 is an exemplary embodiment of the disclosed growing system;

[009] Fig. 2A-2B are exemplary growing pods of the growing system of Fig. 1;

[0010] Figs. 2C is an exemplary container of the growing pods of FIG. 2A;

[0011] Fig. 2D is an exemplary frame of the growing pod of Fig. 2A;

[0012] Figs. 3 A-3B are schematically illustrate the stacking of the growing pods of Fig. 2A;

[0013] Fig. 3C illustrates connecting the growing pod of Fig. 2A to a power supply;

[0014] Figs. 4A-4C are exemplary illustrations of power supply to a growing pod of the current disclosure;

[0015] Figs. 5A-5F are illustrations of different exemplary growing pods and containers that may be used in the disclosed growing system; [0016] Fig. 5G is an illustration of a controller of a growing pod;

[0017] Fig. 6 is an illustration of another exemplary growing pod that may be used in the disclosed growing system;

[0018] Figs. 7A-7B are exemplary lids that may be used in a growing pod of the current disclosure;

[0019] Figs. 8A-8B are exemplary spacers that may be used in a growing pod of the current disclosure;

[0020] Figs. 9A-9B are exemplary grow cups that may be used in a growing pod of the current disclosure;

[0021] Figs. 10A-10B are exemplary light sources that may be used in a growing pod of the current disclosure;

[0022] Fig. 11 illustrates multiple sets of stacked growing pods arranged in a growing facility;

[0023] Fig. 12A-12B illustrates an exemplary rover that may be used in the disclosed growing system;

[0024] Figs. 13-14 are exemplary methods that may be employed in the disclosed growing system;

[0025] Fig. 15 is an exemplary computer numerically controlled machine (CNC) machine that may be used in the disclosed growing system;

[0026] Figs. 16 is an exemplary tool that may be used with the CNC machine of Fig.

15;

[0027] Figs. 17A-17B are an exemplary container and a lid, respectively, that may be used in the disclosed growing system;

[0028] Figs. 18A-18L illustrate exemplary tools that may be used with the CNC machine of Fig. 15;

[0029] Figs. 19A-19B are exemplary illustrations of the CNC machine of Fig. 15 removing and placing a container in a growing pod, in an exemplary embodiment; [0030] Fig. 20 is a schematic illustration of an exemplary growing system of the current disclosure in a greenhouse;

[0031] Figs. 21 A-21G are schematic cross-sectional view of different exemplary growing pods and containers that may be used in the disclosed growing system; and

[0032] Fig. 22 is a schematic illustration of a controller of a growing pod of the current disclosure.

DETAILED DESCRIPTION

[0033] In the following detailed description, reference is made to the accompanying drawings (where like numbers represent like elements), which form a part of this disclosure. The drawings illustrate specific exemplary embodiments of the current disclosure. As would be recognized by a person skilled in the art, other embodiments may be utilized by making variations of the disclosed features without departing from the scope of the disclosure. The following detailed description is, therefore, not to be taken in a limiting sense, and the scope of the current disclosure is defined only by the appended claims. In the following description, numerous specific details are set forth to provide a thorough understanding of the disclosure. However, the disclosure may be practiced without these specific details. In other instances, well-known structures and techniques known to one of ordinary skill in the art have not been shown in detail in order not to obscure aspects of the disclosure.

[0034] The current disclosure describes systems, methods, and devices for an autonomous farming or growing system. Fig. 1 illustrates an exemplary embodiment of the disclosed growing system 1000. With reference to Fig. 1, growing system 1000 may include growing pods 100 on which plants are grown. In some embodiments, these pods 100 may be arranged horizontally (e.g., in rows and columns) on a surface of the growing facility (e.g., floor of a climate controlled indoor growing facility, ground surface of an outdoor facility, etc.). In some embodiments, multiple growing pods 100 may be stacked vertically one on top of another, and multiple stacked pods 100 may be arranged on the surface (see, for example, Figs. 11, 20). In general, any number of pods 100 (e.g., 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, etc.) may be stacked one on top of another. In some embodiments, eight pods 100 may be stacked. The growing pods 100 may be used to grow different types of plants or crops. Different types of pods 100 may be used in growing system 1000. These different types of pods 100 may include, among others, hydroponic pods, aeroponic pods, soil pods, etc. These different pod types may allow different growth methods. For example, hydroponic pods have no soil and the crops grow directly in nutrient-rich solutions, in aeroponic pods the roots of the crops hang in the air and a mist of nutrient-rich solution is sprayed or directed on the roots, soil pods that support soil and an associated irrigation system and are typically used for larger crops that will benefit from a strong root system. The growing pods 100 of growing system 1000 will be described in more detail later.

[0035] The growing system 1000 may also include mobile rovers 200 that are configured to move (i.e., transport) a stack of one or more pods 100 from one location to another. The rovers 200 may include any type of automated (or automatic) guided vehicle (AGV) configured to transport the pods 100. The growing system 1000 may also include one or more robotic forklifts 300. These forklifts 300 may be used in combination with the rovers 200 and may be used to stack, unstack, and shuffle pods 100. For example, a mobile rover 200 may transport a stack of (for example, four) pods 100 from one location of the growing facility to a location proximate a forklift 300. Lifting arms or forks 310 of the forklift 300 may then engage with, support, and lift one or more pods 100 off the stack of pods 100 on the rover 200 and reposition these pods 100. The forklift 300 may include motors 320 operatively coupled to the forks 310 (e.g., via belts, chains,, gears, etc.) and operate to raise and lower the pods 100. Thus, the stacked pods may be unstacked (e.g., using forklift 300) ahead of maintenance or pruning, so that it is easy for an operator (human or robotic) to attend to.

[0036] The growing system 1000 may also include a control system 400 with software that controls the operations of the growing system 1000. The control system 400 may be positioned in the growing facility or may be positioned remote from the growing facility and may (for example, using wired connections and/or wirelessly) directs some or all tasks of the pods 100, rovers 200, and the forklifts 300.

[0037] “Aeroponics” is the process of growing plants in an air or a mist environment without the use of soil or an aggregate medium. In a hydroponic system, plants grow without soil. Instead, they are grown with added nutrients in sand, gravel, or liquid. Aeroponics, a form of hydroponics, that uses no growing medium at all. This is because plants do not require soil to grow, and soil can hinder the specific plant’s growth.

Because all plants need nutrients, the organisms expend valuable energy growing roots to find these nutrients for flower formulation and growth.

[0038] Hydroponic systems, like aeroponics, instead deliver nutrients straight to the source. With a hydroponic growing system, plants are placed in a growing medium, such as, for example, coconut husks, perlite, or clay pebbles. A nutrient-rich solution flows through the airy planting medium and provides food for plant growth. Hydroponic systems provide complete control over nutrient delivery. Plants grown hydroponically have much greater energy efficiency than plants grown in soil. In some embodiments, hydroponic systems also recycle water, which greatly reduces waste. In fact, these soil- free cultivation systems use as little as 10-percent of the amount of water needed by conventional growing methods and are fairly easy to build. Hydroponic gardening uses little to no herbicides or pesticides, and such gardens require little space and are not dependent on growing seasons. Instead, they use artificial light. Because the nutrient solution is passed between plants, it is possible for water-based disease to travel rapidly between them. The aeroponics system does not utilize any growing medium. Plants are suspended in a dark enclosure, while a nutrient-dense solution is sprayed on the roots at certain intervals. Because plant roots are isolated and there is no growing medium, plants that are grown with this suspended, misted system will get maximum nutrient absorption. Aeroponic systems are sensitive and require constant attention to pH and nutrient density ratios.

[0039] An automated guided vehicle or automatic guided vehicle (AGV) is a portable robot that follows along a preselected path on a surface (e.g., marked long lines or wires on the floor) or uses another navigation mechanism (e.g., using radio waves, vision cameras, magnets, lasers, GPS, etc.) to travel along a path on the surface. AGVs are typically used in industrial applications to transport heavy materials around a large industrial building, such as a factory or warehouse. Typical AGVs need some form of external guidance, whether it is permanent wires, magnetic strips or sensors embedded in the floor. This creates a rigid system that is difficult and expensive to adjust as production needs change.

[0040] An “autonomous robot” is a robot that performs behaviors or tasks with a high degree of autonomy (i.e., without external direction or influence). Autonomous robotics is usually considered to be a subfield of artificial intelligence, robotics, and information engineering.

[0041] Internet of Things (IoT) is a system of interrelated computing devices, mechanical and digital machines, objects, animals or people that are provided with unique identifiers (UIDs) and the ability to transfer data over a network without requiring human-to-human or human-to-computer interaction.

[0042] Software as a service (SaaS) is a software licensing and delivery model in which software is licensed on a subscription basis and is centrally hosted. SaaS is sometimes referred to as “on-demand software,” and was formerly referred to as “software plus services” by Microsoft. SaaS is typically accessed by users using a thin client, e.g. via a web browser. SaaS has become a common delivery model for many business applications.

[0043] Fig. 2A illustrates an exemplary growing pod 100. Pods 100 may include a container 10 (e.g., a plastic container) with walls (side walls and bottom wall) 16 that define a root chamber 50 and a nutrient reservoir 60 therein. A lid 12 engages with the peripheral edges of the walls 16 to enclose the root chamber 50 and the nutrient reservoir 60. Since keeping the roots of a plant dark promotes plant growth, in some embodiments, the walls 16 of the container 10 and the lid 12 are configured to keeps the root chamber dark. That is, the walls 16 of the container 10 and/or the lid 12 may be made from a material having a dark color to block sunlight from the root chamber 50 and the nutrient reservoir 60 so that the plants are healthy and not in a stressed state. The lid 12 may be configured to separated and reattached to the container 10. The lid 12 includes multiple openings 14 into the root chamber. Plants are grown in the pods 100 such that the roots of the plants extend into the root chamber through the openings 14 of the lid 12. In some embodiments, grow cups (see Figs. 9A, 9B) on which plants are grown are positioned in the container 10 through the openings. As will be described in more detail later, pods 100 may also include light sources 42 (e.g., see Fig. 10A), nutrient pump(s) 27, sensor(s) 96, 92, controlled s) 26, and electrical cables or wires 25 that provide electricity to power the lights and other electrical components.

[0044] To form a pod 100, a container 10 is removably placed on a frame 20. Figs 2B and 2C illustrate a container 10 and a frame 20, respectively. The frame 20 may be a structure formed by vertical support bars 24A and horizontal support bars 24B that together define a cavity. The container 20 is received in the cavity. Frame 20 may include any number of support bars 24A, 24B. As illustrated in Fig. 2C, the vertical support bars 24A form the feet of the frame 20 that extends between a top and a bottom end. The horizontal support bars 24B connect the vertical support bars 24A and provide rigidity to the frame 20. When the rover 200 moves a stack of one or more pods 100 (referred to herein as a set of stacked pods or a stacked set of pods), the rover 200 may support the frame 20 from below. When the rover 200 lifts the pods, the point of contact and load may be centered to the bottom part of the lowest pod. The horizontal support bars 24B may be positioned and spaced apart to enable the rover 200 to support, lift, and move the set of stacked pods from one location to another in a stable manner.

[0045] With additional reference to Figs. 3A-3C, the pods 100 are stacked such that the bottom end of the frame 20 of the upper pod rests on, and is supported by, the top end of the frame of the lower pod. In general, the frame 20 may be made of any material. It should be noted that the specific structure of the frame 20 illustrated in Fig. 2D is only exemplary, and in general, frame 20 may be any structure that is configured to receive the container 20. In some embodiments, to allow vertical stacking of pods 100, the frame 20 may be made of a metal (e.g., aluminum, steel, etc.). Electrical cables 25 extend though each pod 100. For example, as shown in Figs. 2A and 3A, cables 25 may extend between the top and bottom ends of some or all the vertical support bars 24A of the frame 20. These cables 25 provide power to the electrical components (e.g., lights, pump(s), sensor(s), controller(s), etc.) of pod 100. The cables 25 are connected to electrical connectors 22A, 22B, at both ends of the frame 20. The top ends of the support bars 24A of frame 20 are connected to a first connector 22A and the bottom ends of the support bars 24B are connected to a second connector 22B. When multiple pods 100 are stacked together, the first connectors 22A at the top end of the lower frame 20 (i.e., the frame of the lower pod) engages with the second connectors 22B at the bottom end of the upper frame 20 to direct power between the pods. That is, from the lower pod to the upper pod or vice versa (see Fig. 3B). Thus, the cables 25 of each pod 100 of a stacked set of pods 100 are electrically chained (e.g., connected in series) so that electrical components of all pods 100 of the stacked set are powered.

[0046] External electrical power may be provided to the pods 100 in multiple ways.

For example, power to the stacked set of pods 100 may be provided from the top end or the bottom end of the stacked set of pods. In some embodiments, the second connectors 22B at the bottom end of the lowest frame 20 may engage with power connectors 28 provided on the floor of the growing facility to power the stacked set of pods (see Figs. 3B and 3C). In some embodiments, connectors extending from above (not shown) a stacked set of pods 100 may engage with first connectors 22 A at the top end of the top most frame to power all the pods. In some embodiments, as seen in Figs. 3B and 3C, a flat conductive busbar 30 may be laid on the floor of the growing space and spaced apart power connectors 28 may be positioned on the busbars 30. These power connectors 28 engage with second connectors 22B on feet of each frame 20 to provide power to the pods 100. Once these connectors 22B and busbars 30 engage, a circuit is closed, and the current flows upwards through the lowest pod and daisy-chains all the pods stacked above. As illustrated in Fig. 3C, the spacing between the busbars 30 may match the spacing between the feet (i.e., vertical support bars 24A) of the frame 20. In some embodiments, the busbars 30 will be installed on the floor such that one pair of pod feet receives negative current, and the other set of pod feet receives positive current.

[0047] The connectors 22A, 22B, and 28 may have any configuration. Since power connector 28 engages with second connector 22B, and first connector 22A engages with second connector 22B, when multiple pods 100 are stacked together, connectors 22 A and 22B (and power connector 28 and second connector 22B) will have a complementary shape so that the contact area between the connectors is maximized and resistance to current flow is low. In some embodiments, one of the two connectors that engage with each other may have projection (e.g., male plug) and the other connector may have a cavity (e.g., female plug) that receives and mates with the male projection. In a preferred embodiment, connectors 28 and 22A may be cone or frusto-conically shaped. And, connector 22B may have a correspondingly shaped cavity to receive the cone or frusto- conically shaped connectors 22 A, 28. A cone or frusto-conically shaped connectors may assist in the automated robotic “parking” of the pods 100 in the precise locations matching power connectors 28 fixed to the floor. That is, the power connectors 28 serve as alignment features that assists in accurately positioning the pods 100 on the floor.

[0048] In some embodiments, the busbar 30 and power connectors 28 may be combined, thus shielding the conductive bar and exposing the connectors 28 at predefined distances throughout the bar 30. That is, the power connector 28 and the busbar 30 may be integrally formed or formed as a unitary body. The entire connector 28 may be conductive, thus two connectors 28 connected to two separate conductive bars 30 may be required to provide the negative and positive electric flow to a stacked set of pods. In some embodiments, the lower part of the power connector 28 may be shielded to minimize electric shorts. In some embodiments, a dual conductor cone connector may be used. These connectors can also supply both the positive and negative currents via a pin with the alternative flow located at the center of the connector.

[0049] Thus, in embodiments of the current disclosure, when multiple pods 100 are stacked one on top of another, a pod positioned lower in the stack supplies power to the pod positioned above them. That is, all the pods 100 of a set of stacked pods are collected electrically in series. In some embodiments, this may be achieved by having a correspondingly shaped connector on the support extensions of the pod’s frame as are installed on the floor. In an alternative design, power may be supplied to a pod from a pod above rather than below. For example, a connector that descends from the roof of the growing facility may engage with a connector 22A at the top end of a frame. It is also contemplated that, in some embodiments, power to a pod may be provided from a side of the pod. For example, electric power to a set of stacked pods may be provided by a cable extending from a pod 100 and plugged to a wall socket.

[0050] In addition to power, in some embodiments, connectors 28, 22A, 22B may include or support additional wires (or electrical conductors) to direct other types of signals (e.g., communication signals, etc.) to and from the stacked set of pods. In such embodiments, when two connectors mate or engage, multiple electric contacts in the mating connectors engage to provide power and signal connectivity to the stacked set of pods. These additional conductors may allow the electrical/electronic devices of a pod 100 to operate with a wired communication protocol instead of a wireless communication protocol. The electric controllers (e.g., microcontroller 26 described below) and other electrical devices (pump 27, light sources 42, etc.) on each pod 100 will get both electricity and communication signals over separate wires to signal the activation of relays or send sensory readings.

[0051] Power outages may be a considerable risk in a growing facility, especially when growing with NFT hydroponics or aeroponic pods, as the roots in those pods are not submerged in the solution and will begin to irreversibly dry within just a few hours. To mitigate this risk, in some embodiments, each pod 100 may include an integrated battery backup 21 (see Fig. 2A) that provides power to the pod 100 in the event of a power outage or other power disruption. Additionally or alternatively, a central UPS (uninterruptible power supply) may be electrically coupled to the one or more stacked set of pods to provide continuous power to the pods. A central UPS requires less maintenance over time, however an individual battery backup on each pod 100 has an additional advantage that the pods can communicate wirelessly even when they are not connected to power (for example, when a storage rover 200 is moving them throughout the growing facility).

[0052] As explained previously, set of stacked pods are placed on a floor that includes embedded power connectors 28. The floor may also include navigation guides to assist a rover 200 when moving a set of stacked pods from one location to another in the growing facility. As illustrated in Fig. 4A, in some embodiments, the floor of the growing facility may be formed using (e.g., paved with) prefabricated flooring (e.g., prefabricated floor panels 32 or tiles) that includes embedded power connectors 28, navigation guides 34, and location identifiers 35 (e.g., QR codes). The prefabricated flooring may be flexible or rigid, and multiple panels 32 (or tiles) of prefabricated flooring may connect with each other to form the floor. Each panel 32 may include electric connectors 36 that engage with the connectors 36 of an adjacent panel 32 to transfer electric power (and other signals) from one panel 32 to another. In some embodiments, the prefabricated floor tiles may simply be placed on top of the existing floor. In some embodiments, as illustrated in Figs. 4B and 4C, the floor of the growing facility may include cables 38 with power connectors 28 installed at predefined spacings along the cable. In some embodiments, as shown in Fig. 4B, these cables 38 may be embedded in the floor. In some embodiments, as shown in Fig. 4C, these cables 38 may be laid on the surface of the floor with a ramp like shield 39 (made of plastic or metal) over them. In some embodiments, as shown in Fig. 4C, the shield 39 may include one or more cavities through which the cables 38 extend. In addition to electrically shielding the power cables 38, the height of the shield 39 may also be such that the mobile rovers 200 can move over them. In some embodiments, as illustrated in Fig. 4C, the shield 39 may have multiple cavities that enables conductors to extend therethrough side-by-side with ramp-like features on both sides of the shield. The growing facility floor (panels 32, shield 39, cables 38, etc.) may be able to support the heavy loads of a rover 200 supporting a stack of pods 100. In some embodiments, a storage rover 200 may weigh about 500 Kilograms (Kg). In some embodiments, a heavy pod (e.g., a hydroponic deep water culture or DWC pod) may weigh about 230 Kg, and a light pod (aeroponics or hydroponic nutrient film technique or NFT pod) may weigh about 60 Kg. Thus, with stacked set of eight pods 100, the total load on the floor (and occasionally on the power cables) may be 2300 Kg for DWC or 860 Kg for aeroponic/NFT pods. The prefabricated panels 32 of the growing facility (or the cables extending through the floor of the growing facility) will be configured to support these loads.

[0053] Although in general, the pod 100 may be made of any suitable material, in some embodiments, the main body of the pod 100 may be a plastic container 10 that includes the root chamber 50 and nutrient reservoir 60 (see Fig. 2C). In some embodiments, a food-safe plastic material may be used. In some embodiments, a plastic material having a dark color may be used to block any light from entering either side of the pod. A dark colored container may assist in proper root growth while minimizing the risk of fungi or disease developing in either chamber. The pods 100 (and components of the pods) may be fabricated by any suitable process. In some embodiments, as the pod lids 12 are a separate and interchangeable piece, the container 10 may be manufactured by injection molding or via vacuum forming. The lid 12 may also be manufactured in a similar manner to the containers 10. As explained previously, in some embodiments, the frame 20 may be made of metal. However, when lighter pods are used (such as, for example, hydroponic NFT or aeroponic pods), and the number of pods in a stack is low (e.g., 2-3 pods), the entire pod (i.e., the container 10 and frame 20) may be formed (e.g., integrally formed) from a plastic mold. That is, in some embodiments, the container 10 and the frame 20 may be a unitary body.

[0054] In some embodiments, as illustrated in Fig. 5A and the schematic cross-sectional view of Fig. 21 A, the pod 100 may include a frame 20 (e.g., metal/aluminum frame) lined with lined with a waterproof fabric 40 (e.g., a strong and dark nylon-type fabric) which forms the root chamber 50. Any type of flexible and/or water-proof fabric may be used as fabric 40. A lid 12 may be placed on the fabric to define a container 10A with root chamber 50 and nutrient reservoir 60 therein. This minimizes cost, as only the pod lid 12 need to be plastic molded to relatively precise dimensions. Precise and consistent dimensions may maximize the success rate of the autonomous robots.

[0055] Hydroponics is a method of growing crops without soil. With the hydroponic growing system, plants are placed in a growing medium, such as coconut husks, perlite, or clay pebbles. A nutrient-rich solution flows through the airy planting medium and provides food for plant growth. Hydroponic systems provide complete control over nutrient delivery. Plants grown hydroponically have much greater energy efficiency than those grown in soil. As the nutrient-rich solution is cycled through repeatedly, this method uses as little as 10 percent of the amount of water needed in conventional growing methods. Hydroponics also use little to no herbicides or pesticides as they are less sensitive to diseases as soil systems. A basic hydroponic pod incudes a root chamber 50, a nutrient reservoir 60, a method to oxygenate the nutrients, and a method to deliver them to the roots of the crops.

[0056] Fig. 5B illustrates an exemplary DWC (Deep Water Culture) hydroponic pod 100 (and Fig. 21B illustrates its schematic cross-sectional view). In a DWC hydroponic pod 100, the container 10B that combines the root chamber 50 and the nutrient reservoir 60 so the roots grow directly in the nutrient-rich solution 62 (shown by dashed lines). An air bubbler 64 (or an air stone connected to the microcontroller 26 and an air pump) in the container 10B may operate to oxygenate the nutrient-rich solution 62. The air can be pumped via the air stone or through molded pours at the bottom of the pod. Typically, the main disadvantage of DWC pods is their significant weight, as the entire root chamber 50 is filled with the nutrient mix 62. A considerable advantage is that DWC pods are more resistant to power outages, as the roots are constantly submerged in the solution. While power is required to oxygenate the solution, the roots can easily survive 24+ hours without additional oxygenation.

[0057] Figs. 5C and 21C illustrate a container IOC used with an exemplary NFT (Nutrient Film Technology) pod having a root chamber 50 with a sloped floor 52 to form a nutrient reservoir 60 the bottom of the slope. The sloped floor 52 allows water (or the nutrient solution) to trickle down from the roots to the reservoir 60. A pump 27 controlled by microcontroller 26 pumps the nutrient solution back to the top of the pod to continuously provide nutrients to the roots. In some embodiments, the sloped floor 52 may have an angle or a grade of about 1/100 so that the excess nutrients drip back down to the nutrient reservoir. The drip down to the reservoir and then pumping the mix up to the root chamber also oxygenates the mixture, which promotes healthy growth of the crops. A pod 100 with such a container IOC may be preferred since it provide the lowest overall weight. In some embodiments, to keep the pods working with minimal human maintenance, multiple pumps 27 (e.g., two pumps) may be provided for redundancy. Some embodiments of NFT pods may employ a partially submerged root chamber which the pump(s) constantly cycle and oxygenate. In some embodiments, a dry root chamber (where the nutrient solution is pumped constantly) is used. Partially submerged root chambers add to the overall weight of the pod. However, similar to DWC pods, they are more resilient to power outages, as the roots stay wet even during power outages. Hydroponic pods allows almost full control of the nutrient-rich solution that the crops consume. There are many advantages to this, including customized growth protocols that allow the growth of tastier crops with a longer shelf life, and also stronger crops that grow well in warm/humid climates. This allows the farmer to minimize the farm’s overall energy consumption, as less cooling/dehumidifying will be required due to the custom growth protocol suitable for the warmer climate inside the farm.

[0058] Aeroponics is a method of growing crops in the air with tiny “fog-like” droplets of nutrient-rich solution that is sprayed onto the roots. The roots of the crops are suspended in a dark root chamber, while a nutrient-rich solution is sprayed on the roots at certain intervals. As plant roots are isolated and there is no planting medium, plants that are grown with this suspended, misted system will get maximum nutrient absorption. Research has shown that this is the fastest and most resourceful way to grow many crops.

[0059] Figs. 5D and 2 ID illustrate a container 10D that may be used with an aeroponic pod 100. Container 10D includes a dark root chamber 50, a nutrient reservoir 60, and nozzles 72 or sprinklers to turn the nutrient solution to tiny “fog-like“ droplets. In container 10D, the nutrient reservoir 60 and root chamber 50 are combined. The top of the container 10D is used as the root chamber 50, in the middle are the nozzles 72 which spray the roots and the bottom of the container is the nutrient reservoir 60 (the nutrients fall to the bottom, from where they are pumped back to spray). The size of the droplets may influence the growth rate and nutrient absorption of the roots. A mist of the nutrient solution can be achieved through a high-pressure pump 27 that pushes the nutrient solution out through nozzles 72 with tiny holes of the correct droplet dimension. Alternatively (or additionally), as shown in Figs. 5E and 2 IE, an aeroponic pod may use a container 10E with electric piezo drivers 74 or ionizers placed within the nutrient mixture. When the ionizers are powered, they vibrate at high frequencies (e.g., 1 Mhz, 2 MHz, 3 MHz, etc.) to turn some of the mixture to small mist droplets. Experiments using 1.7Mhz and 2.4Mhz disk ionizers were successfully used for growing lettuce aeroponically. Other frequencies may also be used as they change the droplet size.

[0060] NASA research has shown that aeroponics is the fastest growing method, as the roots need to make the least amount of effort to consume the nutrient solution in comparison to soil or hydroponics. Furthermore, aeroponic pods are extremely lightweight due to the small amount of nutrient solution required as the nutrient is turned into a mist unlike hydroponic methods. The main disadvantage of aeroponic pods is that they are extremely vulnerable to power outages as, without the pumps/ionizers, the roots will begin drying (from which there is no way to resuscitate them) within roughly two hours. Therefore, to keep the pods working with minimal human maintenance, multiple pumps 27 (e.g., two pumps) may be used for redundancy. [0061] Soil pods may be used for larger plants with strong roots or heavier crops. Figs. 5F and 2 IF illustrate a container 10F that may be used with a soil pod. Container 10F may include a dark root chamber 50, filled with soil, a small nutrient reservoir 60, and a pump 26 that provides precise amounts of the nutrient mix and irrigates the soil. Unlike hydroponic and aeroponic pods, the nutrient mix are to be pumped to the top of the pods, preferably to the lids 12 (see Fig. 7B) so that the mix can be dropped precisely above the crop to minimize water waste. A clear disadvantage of soil pods is that they are very weighty compared to hydroponic NFT or aeroponic pods. A key advantage of soil pods is that almost all types of crops can be grown in this medium. Soil pods are also the least vulnerable to power outages, and crops can continue to grow for at least three days until power is restored and additional irrigation is required.

[0062] Figs. 6 and 21G illustrate an alternative embodiment of stacking pods for a vertical farm layout that can be achieved without centrally controlled or electrically powered pods 100 (described, for example, with reference to Figs. 2A-3C). In this embodiment, the floor installation may not provide electricity. Instead, as shown in Fig. 6, a pressurized nutrient-rich solution is provided to the containers 10 of pods 100' through conduits 35 in the frame 20. A nutrient-rich solution may be directed into the conduits 35 of the pods 100' from a central tank (for example, via fluid conduits that extend through the growing facility similar to that described with reference to Figs. 4A- 4C). The frame 20 may also include conduits to drain the solution back from the pods 100' to a central tank. The central tank contains a nutrient-rich solution that is used for all the pods 100', thus eliminating the possibility of using different grow protocols for different pods. In some embodiments, as with pods 100, multiple pods 100' may be stacked with conduits 35 of a lower frame fluidly coupled to the conduits 35 of the frame above via fluid couplings 35 A, 35B. As described with reference to electrical connectors 22 A and 22B of pod 100, the fluid couplings 35 A and 25B on the top and bottom ends of pod 100' may have a complementary shape such that coupling 35 A at the top end of the lower pod 100' engages with coupling 35B at the bottom end of the upper pod 100' to transfer fluid between these pods 100'. As in electrically powered pods 100, pods 100' may also include artificial lights, pumps 27, and microcontrollers 26 to control the pod 100'. [0063] In some embodiments, pods 100' may not be stacked as high as electrically powered pods 100, and pods 100' may not include artificial lighting. In such embodiments, the pods 100' placed on top may shade the pods 100' below it. In some embodiments, common artificial lights may be installed in the growing space, or the growing space may be configured allow natural light to provide light to the pods 100'.

[0064] Although pods 100' may, in general, use any of the previously described type of container, in some embodiments of powerless pods 100' may include a hydroponic NFT container is recommended. A fluid connector on the foot of the pod 100' connects to a floor connector that provides the pressurized nutrient-rich solution. A connector on the other side of the pod (e.g., on another foot) connects to a separate connector that provides a line for draining the pumps. The drain line may be under negative pressure to assist with draining the pods. When the central pump is activated, the nutrient-rich solution is pumped from the central tank throughout the farm’s pods. These pods are designed so that the root chamber is at an angled slope, which causes the solution to gravitate towards the second side of the pod, where the drain line is connected. This method allows nutrients to constantly cycle and flow throughout the root chamber without needing a pump for each pod. This drastically reduces the cost and complexity of each pod. Unlike electrically powered pods 100, the male connectors (installed on the floor of the growing space) and the top of the support stand of each pod may include a valve that stops the fluid flow if no pod is mounted above it.

[0065] As described below, smart software (based, for example, on artificial intelligence) may be used to monitor all of the sensory and climate readings throughout the growing space (or farm) and the stacks of pods 100, 100' therein and make active changes to accommodate the climate. In some embodiments, this software may be included in control system 400 (see Fig. 1). These changes may include some or all of: temperature/humidity change via an air-conditioning-like unit; changing the carbon dioxide (CO2) content in the air; changing the nutrient-rich solution mixture to a different recipe; changing the artificial lighting schedule/intensity or spectrum; and changing the timing of the pumps that control the nutrient reservoir. The artificial intelligence model requires a very large amount of data, tested with different climate and growth protocols so that it can evolve and make better autonomous decisions for commercial farms. [0066] To support the creation of such a large dataset with variations of climate and nutrients, an enclosed, fully controllable R&D pods may be used. The R&D pods may be similar to the pods 100, 100' described above (and can support any grow method, hydroponics, aeroponics or soil), with some or all of the following changes: these pods are fully enclosed so that light does not enter the pod, and the crop is only influenced by the artificial lighting installed on the top of each pod; climate-controlled air is pumped into the pod, allowing the software to control the temperature and humidity within the pod precisely; CO2 is similarly pushed to the pod for precision control of the CO2 levels within the pod. These pods may be used for the research and development of growth protocols suitable for different climates, crops, and throughout the lifecycle of each crop.

[0067] Before crops (or grow cups) are placed in the pods, and an interchangeable lid 12 may be put on top of each pod. Aeroponic and hydroponic crops are placed into grow cups (described below) that are then placed on the lids 12. The crops or plants are planted in the grow cups and the grow cups are positioned in the container of the pod.

The lids 12 of the containers have openings 14 that are slightly smaller than the grow cups and prevent the cups from falling into the root chamber of the container. In embodiments, where soil pods (see Fig. 5F) are used, grow cups may not be used to hold the crops. Instead, the crops are planted directly into the soil of the pod. For these pods, the lid is used to deliver the nutrient mix to each crop individually. A key advantage to grow cups and hydroponic/aeroponic pods is that the crops can be lifted and moved, manually or robotically, between different pods. This is especially beneficial to maximize the crop density of multiple pods. For example, for the first ~14 days of growth from seed, the seedlings are still very small and require a minuscule amount of space. On a 1 sqm. pod, at least 100 seedlings can be placed. As the seedlings grow, they need more and more space. Lettuce, for example, can be placed at a maximum density of 30 crops per 1 sqm. pod during the final week of growth.

[0068] Traditional systems require the seedlings to be planted in a way that they have enough free space around them for the final period when they are largest. In this system, however, plants can be replanted (picked up from one pod to another) throughout their lifecycle. By using different lids 12, with a different number of openings 14 and free space around each hole, the number of crops grown in any given space may be maximized. Additionally, the lids 12 block light from above from entering the root chamber and stressing the plants. Fig. 7A illustrates an exemplary lid 12 with openings 14. The perimeter of lid 12 may have a lip region that engages with the perimeter of a container 10 to removably attach the lid 12 on the container 10. It should be noted that, in the discussion below, container 10 generally refers to all the previously described containers 10A-10F. To further maximize the absorption rate of the light, which, in turn, can lower the overall energy consumption, the top side 13A of the lid 12 may be coated or covered with a reflective covering that will illuminate the lower parts of the crop. Lid 12 may include any number of openings 14 (10, 20, 40, 60, etc.) of any size. Although the openings 14 may be circular 14 in some embodiments, in general, the openings 14 may have any configuration (circular, square, rectangular, etc.). The size and density of the openings 14 may be selected based on the crops grown. In embodiments, where grow cups are used, the size (diameter, width, etc.) of the openings 14 may be slightly smaller than the size of the grow cups such that the rim of an opening 14 supports the grow cup positioned in the opening 14.

[0069] In some embodiments, as shown in Fig. 7B, the lid 12 (e.g., top side of lid) may include tubes or conduits 15 to provide water (or a nutrient-rich solution) to the plants grown in the container. The conduits may direct water to the plants via the openings 14. Water pumped from the nutrient reservoir (described previously) may be directed to the plants through these conduits 15. Each plant may receive the desired amount of nutrient via the conduits 15. In some embodiments, the conduits 15 may be molded together with the lid 12. That is, the lid 12 and conduits 15 are not separate components, instead, they are integrally formed. Tubes or conduits 15 may also be used to push air to the air bubblers (see Figs. 21 A, 21B) and for pumping nutrient from the reservoir 60 to sprinklers 72 in Aeroponic pods (see Fig. 2 ID) or directly to the roots in NFT pods (see Fig. 21C).

[0070] As can be seen in Fig. 3B, the size of the frame 20 typically controls the spacing between the pods 10 in a set of stacked pods. In some cases, additional height may be required between the pods 100 of a stacked set. For example, as big crops grow taller, additional vertical space may be required. To achieve this, spacers may be used to increase the vertical distance between the pods while still allowing most of the light to pass through them. For example, lettuce and green herbs such as basil require roughly 60 cms between the crop and the artificial lighting (30 cm of space is used for the growth of the crop, and the additional 30 cm is required for adequate spacing between the light source and the crop). Tomato plants, for example, will require different distances and heights between the pod and the light source. For the first two months, the same 60 cm as lettuce plants is sufficient. Spacers will then be required thereafter. Figs. 8A illustrates an exemplary spacer 70A that may be positioned between two adjacent pods of a stacked set (or below the lowest pod of the stacked set) to increase the vertical spacing between them. Similar to frame 10 of the pods, the support members of the spacer 70A may also include cables 25 and connectors 22A, 22B that directs electrical power between the pods on either side of the spacer 70A (or from the floor to the lowest pod).

[0071] Some crops, such as young seedlings or medicinal crops, may occasionally benefit from total and complete darkness. For this purpose, a shading spacer 70B as shown in Fig. 8B may be used. The shading spacer 70B blocks light from all directions can be robotically placed above the pod. A fabric 76 may extend over the frame of the spacer 70B to block light.

[0072] Most pod types (except for powerless pods) have an independent nutrient reservoir. By having an individual reservoir for each pod, the system can use different nutrient recipes for varying crops at different stages. The root chamber is funneled back to the nutrient reservoir, thus allowing the reuse and cycling of the same solution with minimal resource waste. When the nutrient solution is pumped/ionized back to the root chamber, it is mixed with air and oxygenated, which is essential for optimal crop growth. The nutrient chamber can be molded as part of the growing pod.

[0073] The temperature of the nutrient solution can either be maintained by full climate control systems of the ambient temperature in the farm or via a local cooling system (such as adding a fan which cools the tubes, and nutrients, as they’re pumped from the reservoir to the root chamber)

[0074] Figs. 9A and 9B illustrate exemplary grow cups 80A, 80B that may be used in the pods. Grow cups 80 A, 80B are semi-open containers or cups (e.g., made of plastic, etc.) that hold the crop and its growing medium (such as coconut husks, perlite, clay pebbles, etc.). These cups 80A, 80B are then placed into the openings 14 on the lids 12 of the growing pods. To robotically lift and move the grow cups, as illustrated in Fig. 9A, some preferred embodiments of grow cups 80A includes graspers 82A that robots can easily grab and securely move. Although any number of graspers 82A may be provided in a grow cup 80A and they may be positioned at any location on the grow cup 80A, some embodiments include two graspers 82A positioned diametrically opposite each other. It should be noted that, the graspers 82A may have any configuration that enables a robot to grab and hold. In some embodiments, the graspers 82A may be shaped like hooks. Alternatively, or additionally, in some embodiments, as illustrated in Fig. 9B, a ring 82B (e.g., a metal ring) may be positioned around the opening of the grow cup 80B . In such embodiments, an electromagnet (or a suction device) may engage with the ring 82B to move the grow cup 80B.

[0075] Each pod 100, 100' may have a microcontroller 26 configured to control some or all the electrical components of that pod 100. For example, microcontroller 26 of each pod 100 may control the lights of the pod, activate the pumps and/or ionizers of the pod 100 (or associated with the pod) to deliver nutrients to the roots of the plants in the pod, etc. The microcontroller 26 may be controlled by control system 400. Figs. 5G and 22 illustrate the details of the microcontroller 26 of a pod 100 in one exemplary embodiment. As illustrated in these figures, microcontroller 26 may be operatively coupled to light sensor 96. The microcontroller 26 may control the light sources 42 using a light controller 95. The light sensor 96 may face the light sources 42 of the pod 100 above. Microcontroller 26 may also be operatively coupled to the pump 27 of the pod 100. Microcontroller 26 may also include power relays 91 or switches that open and close the electrical circuit of the pod 100. In some embodiments, the microcontroller 26 may be remotely controlled by control system 400 by local or cloud software (i.e., via the internet). The microcontroller 26 may connect wirelessly to the local network via WiFi, RF, LoRA, or other wireless protocols. In some cases, multiple pods 100 (adjacent pods in a stack, all pods in a stack, etc.) may be controlled by the same microcontroller 26. In some embodiments, in addition to the microcontroller 26, the controller may include: Pump (None - for Aeroponic pods with Ionizers, Air pump - for DWC pods, Water pump - for soil pods, NFT pods, High pressure water pump - for Aeroponic pods with sprinklers); Sensors (some or all of light sensor, CO2 sensor, sensor to detect climate outside the pod, sensor to detect climate within the pod (for Aeroponics), water level sensor for the nutrient reservoir (for all types), soil humidity sensor; and Relays (control the on/off of other components). The relays may control artificial lighting below the pod (possible for all), Ionizers for Aeroponics. The pump may also be controlled via a relay.

[0076] Microcontroller may include printed circuit board and/or electronic/electric components that are configured to control the pod 100. Since such microcontrollers are known to persons of skilled in the art, it is not described in detail herein.

[0077] Each pod 100 may include light sources 42 (described below) and/or nutrient/irrigation pumps 27 controlled by the microcontroller 26. The electricity to power these electronic or electrical devices may be provided via cables 25 extending through the frame 20 of the pod 100 (see Fig. 2A). When the pods 100 are stacked on top of each other (see, for example, Figs. 3 A and 3B), the electricity is chained vertically up through the stack through cables 25 extending through each pod 100 (e.g., through support bars 24A). This greatly simplifies the power infrastructure required, as power connectors 28 on the floor are all that is required to power the vertical stacks of pods. In some embodiments, the pods 100 are powered via a low-voltage DC current, as this is sufficient for the small lighting sources and pumps. Using low voltage DC, may also minimize risk (e.g., the risk of electrocution).

[0078] The microcontroller 26 of each pod is operatively coupled with sensors for better monitoring and optimization of the farm and crops in the pods. Any number and type of sensors may be used. In some embodiments, these sensors may include a water level sensor 92 (see Figs. 2C, 5F) in the nutrient reservoir 60 so that the system is aware when more nutrient solution is needed. The growth chamber 50 (e.g., of soil pods) may also include a soil humidity sensor 94 that allows automation of the irrigation based on the current status of the soil (see Fig. 5F). In some embodiments, one or more light sensors 96 may be provided to monitor the actual light captured by the pods and automate based on when and how much artificial lighting is required (see Fig. 5F) . In some embodiments, the light sensor 96 may monitor the light spectrums for where photosynthesis occurs, such as UVA, UVB, and UVI. In some embodiments, light intensity sensors (instead of UV-specific spectrum sensors) may be used, for example, to reduce cost. These sensors measure the intensity of light and do not differentiate between the spectrums of light (only some spectrums cause photosynthesis). However, especially in indoor installations that rely fully on the light sources 42 of the pods 100, it is possible to estimate the exposure of the relevant light spectrums precisely. Aeroponic pods (see Figs. 5D and 5E) may also use a humidity sensor 98A placed within the root chamber 50. The humidity sensor 98A may allow automation of when the mist pump/ionizers should be activated. In some embodiments, CO2 sensors 97 may be provided to allow automation of CO2 delivery (see Fig. 5E). In some embodiments, temperature and humidity sensors 99, 98B may also be positioned outside some or all of each pod to determine the ambient conditions around the pod. This information from these sensors may be aggregated to make climate control automation, such as air conditioning or humidifiers. It should be noted that, although the sensors are indicated as being positioned at specific locations of a pod, this is only exemplary. In general, these sensors may be positioned at any suitable location.

[0079] Artificial lighting is required for all indoor farms and also when the pods 100 are stacked are arranged in controlled climate greenhouses. Fig. 20 illustrates an exemplary greenhouse where multiple sets of stacked pods 100 are arranged (e.g., in multiple rows and columns) on the floor. Since greenhouses are semi-transparent, some sunlight does enter. However, as the pods are stacked, the lower pods will be blocked from light due to the pods above them. There are two main approaches for the location of the artificial lights. In some embodiments, as illustrated in Fig 10 A, one or more light sources 42 may be provided below a pod 100 to provide light for the pod below it. These light sources 42 may be positioned below the frame 20 of a pod 100. Although this is a simple solution, it has the disadvantage of no light for the top-most pod and unusable light sources 42 on the bottom of the bottom-most pod. In some embodiments, as illustrated in Fig. 10B, a lighting spacer 70C may be used. Lighting spacer 70C may function in a manner similar to spacers 70A, 70B described with reference to Figs. 8A and 8B. Lighting spacer 70C may include light sources 42 installed below the spacer 70C. A controller 26 with a light sensor 96 and relays 91 may also be attached to spacer 70C. When needed, the lighting spacer 70C may be positioned above the top-most pod or between two pods. As illustrated in Fig. 10B, similar to spacer 70A (of Fig. 8A), lighting spacer 70C may also include a connectors 22A and 22B at its top and bottom ends and cables 25 that extend therebetween. The light sources 42 of spacer 70C may be powered by the cable 25 that extends therethrough. These lighting spacers 70C are modular can be placed between the growing pods and deliver light to the pods as required. The light sources 42 (of Figs.

10A and 10B) may be an LED strip or any other type of light emitting device. In some embodiments, light sources 42 may include spectrum-specific LEDs that are both energy efficient, driven by DC power, and light mainly in the correct photosynthesis spectrums. Alternatively, high-pressure sodium lights can also be used.

[0080] With reference to Fig. 1, one key benefit of prefabricated stackable pods 100 with consistent dimensions is that they can easily be stored, moved, packed, and unpacked with the coordinated movements of two simple robots. These robots are a storage rover 200 used to move towers of one or more stacked pods 100 and a forklift 300 used to stack and unstack the towers of pods 100. Unlike vertical farms that require expensive infrastructure like support frames or racks and pipes, the currently disclosed system allows high densities of pods 100 in all dimensions (horizontal and vertical) with minimal setup. The disclosed system only requires a flat floor (e.g., a concrete floor) for the rovers 200 to navigate around easily with power connectors 28 installed previously described. The disclosed system is also fully flexible, meaning more pods 100, rovers 200, and forklifts 300 can be added to increase the floor space used or the height of the stacked pods 100. The disclosed system can also accommodate non-typical areas that require a combination of heights and different locations for the parked pods. The disclosed system can also be used to position pods 100 in different spaces, with rovers 200 circulating as needed. For example, if there is a limited area with natural light and a larger indoor space. The control system 400 chooses which pods will most benefit from the natural light, based on the crop and stage of growth. This minimizes the need for artificial lighting. The rovers 200 can then circulate those pods to the lit area and back. This can be similarly used to transfer crops between areas with different controlled climates, such as when flowering crops need distinct temperatures or humidity levels to seedlings. [0081] The disclosed system is configured to enable robotic pruning and maintenance but is also far more comfortable for human farmers’ operations. Instead of walking and climbing to attend to crops, those pods 100 requiring maintenance can automatically be delivered to the farmer (using rovers 200), who can remain in one place. Not only is this more comfortable, but as humans do not enter the grow space, it is possible to place pods at a higher density (as illustrated in Fig. 11) and with climates that are beneficial for the crops but not friendly to humans (high levels of CO2, etc.). For example, as illustrated in Fig. 11, since the stacks are packed closely together, the pods of the middle stack are not accessible to humans. However, in the currently disclosed system, a rover 200 can move the stacks around and access any stack easily.

[0082] Different types of storage rovers 200, also known as AMR (Autonomous Mobile Robot) or AGV (Autonomous Guided Vehicle), may be used in the disclosed system.

Fig. 12A illustrates an exemplary storage rover 200. Rover 200 includes a driving system 220 housed in a casing 210 and a lifting jack 260 that extends out of the casing 210. The driving system 220 includes driving wheels 222 (two or four wheels), an electric motor 224, a battery system 226, and a wireless microcontroller 228 that guides the movement of the rover 200. The electric lifting jack 260 coupled to the motor 224 extends out of the casing and is used to lift a stack of pods from the ground and move it to other locations carried on top of the rover 200. Rover 200 may include sensors 240 that assist the rover 200 in its functions. Although not a requirement, in some embodiments, the storage rovers 200 may be smaller than the pods 100 themselves (see Fig. 2D) so that a rover 200 can slide under a set of stacked pads 100 and lift them with taking up any space outside the area of the pods 100. This will reduce the maneuvering space required around the pod. With reference to Fig. 12B, although not a requirement, in some embodiments, the pods 100 and rovers 200 may both be substantially square-shaped, as this allows the rovers 200 to turn in place and easily access the different parking positions with minimal clearance space. In contrast, robotic storage systems designed to also be used by humans - such as eCommerce warehouses - prefer a rectangular rover and shelving as this is easier to reach within the shelving when facing the long side of the rectangle. However, this is not an issue for growing pods, as they are intended to be attended to robotically from above. [0083] Sensors 240 enable the rover 200 to function efficiently (e.g., navigate the farm successfully). Navigation can be achieved by installing magnetic strips on the floor that serve as guide lanes for the rovers 200 to move on. A sensor that identifies the magnetic strips may be used to keep the rover 200 on track. Alternatively, images or radar readings obtained from cameras, LIDAR, or other radar sensors installed on the rover 200 can be used to map and navigate the area. Alternatively, wireless beacons can be installed in set locations throughout the farm, and a sensor can triangulate its precise location based on multiple beacons’ signals. Cameras installed throughout the farm can also assist in identifying the rover and its position, which can be relayed through the central software. Climate sensors 242 (such as temperature, CO2, humidity, etc.) may be placed on the rovers 200 to autonomously sample the farm’s entire climate. As pods 100 must be placed robotically with a high level of accuracy due to the power connectors 28 being fixed to the floor, a camera, facing downwards to the floor, may be used to identify markings (e.g., navigation guides 34 and location identifiers 35 of Fig. 4A) on the floor. The driving system 220 may include either two or four wheels 222. By driving the wheels 222 in different directions, the rover 200 can turn in place.

[0084] An electric jack 260 is installed in the center of the rover 200. When the jack 260 is folded or lowered, the total height of the rover 200 is smaller than the height of the support feet of the frame 20 of the pod 100 (see Fig. 2D). This is so the rover 200 can slide underneath the pod 100. Once the jack 260 is activated (raised, lifted, etc.), the total height of the rover 200 becomes higher than the height of the pod’s feet and the rover 200 lifts the pods 100 (see Fig. 12B). In essence, the rover 200 is pushing and holding the rack of pods 100 in the air. This allows the rover 200 to move the stacked set of pods 100 simply by driving with the jack 260 in the activated position with the pods 100 held above it. Once the rover 200 has arrived at its precise destination, the jack 260 is lowered and the stacked set of pods is placed on the floor. The rover 200 is then free to drive away. The jack 260 may have different embodiments. In some preferred embodiments, an electric motor drives a screw connected to the lifting platform - up and down. A gear system of the jack 260 may support the weight of the set of stacked pods.

[0085] The forklift 300 enables robotic stacking, unstacking, and shuffling of pods. These actions will be undertaken multiple times throughout the growth cycle of the crops. For example: (A) Human or robotic tasks - these include cleaning, seeding, replanting, testing nutrients, refilling the reservoir, pruning, harvesting, etc. and are most efficiently done when the stacked set of pods 100 are unpacked and presented as a single pod 100. For example, this allows the human farmer to perform all of his or her tasks on multiple pods 100 without even moving, as the pods 100 are brought to him/her. Additionally, this simplifies the required robotic freedom of movement considerably and is one of the reasons that the disclosed system uses CNC-like robots. These robots are fixed and stationary for most farming tasks. If the work plane we not fixed and horizontal, such as in vertical towers, a complex robotic arm would be required to reach all required locations. (B) Maximizing natural and artificial light - In a preferred embodiment, the entire grow area is located in a climate-controlled greenhouse, which is semi-transparent so that natural sunlight can pass through to the pods 100. Sunlight is important to lower the farm’s energy consumption by avoiding reliance on artificial lights. Artificial lights are still required, especially when the farm is stacked with multiple levels of pods, as the lower pods will be constantly blocked from the sunlight by the other pods on top and around it. However, by shuffling the location of the pods throughout the day, for example, from a low, shaded location to the top of the stack with direct sunlight, the system can better distribute the light to the pods that are currently in a stage where the light will be most efficient. (C) Optimizing yield - robotic forklifts 300 may be required for the autonomous stacking and unstacking of growing pods 100 as described below. These tasks may be performed by a wirelessly controlled stationary robotic forklift 300, in sync with one or more storage rovers 200. In some embodiments, a forklift may be incorporated on the storage rover 200. However, since the total footprint of the rover 200 may be larger than the pods 100 in such embodiments. A larger rover 200 may reduce the possible packing density of pods 100 as larger pathways and maneuvering space will be required.

[0086] To pack and unpack the pods, a series of synchronized movements by the storage rover(s) 200 and forklift 300 may be performed. Fig. 13 is a flow chart that outlines an exemplary process 2000 for a single rover 200 and forklift 300 to stack three pods 100 to create a three-pod vertical stack. In the description below, reference will also be made to Fig. 1. A rover 200 drives below a pod 100 (pod #1) (step 2010) and raises the pod 100 (step 2012) using its jack 260. A forklift 200 gets ready with its forks 310 positioned ready to engage with the pod 100 (the lower position shown in Fig. 1 called herein as position #1) (step 2014). The rover 200 drives with pod #1 to the forklift 300 (step 2016). In this position, the forks 310 of the forklift 300 may be positioned to support the frame 20 of pod #1 and lift the pod. The forklift 300 moves its forks 310 up to position #2 (the higher position shown in Fig. 1) holding pod #1 in the air (step 2018). Rover 200 then drives and picks up another pod 100 (pod #2) (step 2020). The rover 200 drives with pod #2 to the forklift 200 (step 2022). The forklift 300 moves forks 310 down (step 2024). The rover 200 moves forward with both pod #1 and pod #2 (step 2026). The forklift 300 moves forks 310 down to position #1 (step 2028). The rover 200 drives back to the forklift 300 (step 2030). The forklift 300 moves its forks 310 up to position #2 holding pod #1 and pod #2 in the air (step 2032). The rover 200 drives and picks up another pod, pod #3 (step 2034). The rover 200 drives with pod #3 to the forklift 300 (step 2036). The forklift 300 moves forks 310 down so all three pods 100 are stacked on the rover 200 (step 2038). The rover then drives away with a set of three stacked pods 100 (step 2040). Exemplary process 2000 may be repeated to stack more pods 100 and create higher vertical stacks. Multiple rovers can be used to accelerate the procedure by using multiple robots in parallel for different tasks.

[0087] Fig. 14 illustrates an exemplary process 2100 that may be followed to unstack three pods 100 so each pod is left without any pods above it. A rover 200 drives beneath 3 stacked pods (pod #1 below pod#2 and pod #3 above pod #2) (step 2110) and raises the set of stacked pods (step 2112). Forklift 300 gets ready with forks 310 positioned to engage with pod #2 (position #2) (step 2114). The rover 200 drives with the stack of pods to forklift 300 (step 2116). In this position, the forks 310 are positioned between pod #1 and pod #2. The forklift 300 move forks 310 up to lift pod #2 and #3 in the air (step 2118). The rover 200 drives away with pod #1 and parks it elsewhere (step 2120). The rover 200, without any pod on it, drives back to the forklift 300 (step 2122). The forklift 300 moves its forks 310 down (step 2124) such that both the pods rest on the rover 200. The rover 200 moves forward with both pod #2 and pod #3 (step 2126). The forklift 300 moves its forks 310 down to a position to engage with pod #3 (position #1) (step 2128). The rover 200 with pods #2 and #3 positioned on it drives back to the forklift (step 2130). In this position, the forks 310 are positioned between pods #2 and #3. The forklift 300 moves forks 310 up to lift and hold pod #3 in the air (step 2132).

The rover 200 drives away with pod #2 and parks it elsewhere (step 2134). The rover 200, without any pod, drives back to the forklift 300 (step 2136). The forklift 300 moves its forks 310 down so pod #3 is placed on the rover 200 (step 2138). All pods are now unstacked (step 2140). The process may be repeated to shuffle the pods. This procedure can be repeated to unstack higher vertical stacks with more than three pods. Multiple rovers 200 can be used to accelerate the procedure by using multiple robots in parallel for different tasks. Process 2100 in combination with process 2000 allows shuffling of the pods 100 within a set of stacked pods, for example, for better light distribution.

[0088] As the pods 100 are robotically stored and collected, no humans need to enter the growing space unless for maintenance. By not having humans in the grow space, multiple advantages are achieved. Namely, less/no need for human-specific robot sensors. In some embodiments, some or all robotic warehouse rovers 200 will have delicate, fast sensors that can identify humans so that they do not accidentally run into them. By simply stopping the driving of the rovers 200 when humans enter the farm, the price of each rover 200 can be reduced as human-specific robot sensors are not required; Additionally, CO2 enrichment can be achieved by not having humans in the grow space. While dangerous for humans, plants thrive on high levels of CO2 in the air. As all human-related tasks are performed outside of the controlled grow area (to which the pods 100 are brought robotically), the crops can be grown in ambient with high C02 levels (e.g., CO2 levels that are dangerous for humans). Further, by not having humans in the grow space, the grow space can have minimal wasted space or pathways (e.g., just enough to move the towers of stacked pods, not enough to be human-friendly). The stackable pod concept of the current disclosure allows the growth of crops at unparalleled densities compared to existing solutions and promotes yield optimization.

[0089] To optimize the overall yield of the farm, a combination of methods may be used. Parking pods 100 in inaccessible areas. For example, as the software is aware when maintenance will be required for each pod 100, it can simulate and decide which pods 100 can be parked in a way that they are not conveniently accessible until the time of maintenance arrives. This is possible due to each pod’s sensors, which allow remote monitoring. Shuffling pods to maximize natural light absorption. Optimizing the absorption of natural light both lowers the energy consumption of the farm and also speeds up the growth speed, as natural sunlight is (usually) far more effective for photosynthesis than artificial lighting. With the rovers 200 and forklifts 300, the control system 400 can robotically shuffle the pods 100 (within a single stacked set or between multiple stacked sets) so that the natural light is better distributed between the pods 100 based on the crop, stage, and current light requirements. Shuffling pods to maximize artificial light absorption. Similar to sunlight, pods 100 with artificial lights can be grouped so that the ambient light given off by the activated pods 100 is absorbed by other pods that need light. Crop seedlings, for example, require less light than flowering crops and thus they may be grouped at a distance from each other. Quickly remove sick crops. The control system 400 may identify pods with potential contaminants (based, for example, on sensors in the pods and/or the grow space). And, when it does, the rovers 200 can quickly fetch and remove/check those specific pods 100. This can minimize the chance of airborne infection to the rest of the pods on the farm.

[0090] The current disclosure also promotes or enables robotic precision farming. The disclosed robotic stacked pod growing system is beneficial for both human-operated farms and fully robotic and autonomous farms. Similar to human farmers, who can stay in place, and the pods 100 will be brought to them, precision robotics can also stay in place. This vastly simplifies the complexity and cost of precision-farming robotics.

[0091] The robotics for a fully autonomous farm in an exemplary embodiment is described below. In a pod-based farm system, tasks depend on the crop grown. Routine tasks include: cleaning, adding soil; preparing grow cups; placing pod lids; seeding; photo analysis; analyzing the roots; replanting; watering; testing the pod’s soil; testing, adjusting and managing the nutrient reservoir; pruning; trellising; harvesting; and packaging. To support so many different types of tasks, affordable automation does not seem economically viable with complex robotic arms, especially if installed on robots that move around the farm. While these designs can technically perform most of the required tasks, they will be both extremely complex and expensive, and also decrease the density of the farm due to the pathways required for the robots to move around the farm. (unlike the disclosed storage rovers 200 that can fit under the pods 100, thus requiring no additional moving space).

[0092] When using stacked pods, tasks are all performed on the same horizontal plane of the lid of the pod (from the top, looking down). This significantly simplifies the required freedom of movement and the complexity of the precision farming robotic mechanics. This is also why, in some embodiments, the pods are designed so that their nutrient reservoir is accessible from above.

[0093] The disclosed farming system may use one or more CNC machines 500 (Computer Numerical Control) to assist in various tasks (see Fig. 20). A CNC machine is a motorized maneuverable tool (on a motorized maneuverable platform) which are both controlled by a computer (e.g., control system 400) in accordance with instructions (e.g., preprogrammed instructions) from a user. It should be noted that although referred to as CNC machines, any type of machine that performs the functions of the described machine in an automated manner can be used. With reference to Fig. 15, any known type of CNC machine 500 (e.g., FarmBot) may be used in the current disclosure. In the discussion below, CNC machine 500 may also be referred to as a CNC robot. In contrast with known uses of CNC machines in farming, the current disclosure uses a combination of CNC machines 500 and stackable pods 100. As evident in Fig. 15, in the disclosed system, the stackable pods 100 are attended to by the CNC machine 500 from above. In some embodiments, the CNC machine 500 tends to the pods 100 only from above. A CNC machine 500 may be used to place containers 10 on the frames 20 of a pod 100. As illustrated in Figs. 19A and 19B. a CNC machine 500 may grab a container 10 from the top, lift the container 10 off its frame 20 and place the container 10 on the frame 20 of another pod 100. Handling a pod 100 only from the top enables the pods 100 to be attended to by CNC machines, which are much cheaper. CNC machines, often used for industrial manufacturing, are (typically) table-like stationary machines that can move in the X/Y/Z-axis of the fixed space below them. These machines are limited in movement compared to robotic arms but are far simpler and cheaper due to their mechanical design. Because of the large size of typical farms and the high number of growing pods, several robots will be required. This further highlights the importance of minimizing the complexity and costs related to the robotic solution. [0094] Precision robotics, not only for farming tasks, must be well -calibrated to the exact dimensions of the workspace. Achieving this level of calibration when the robot is constantly moving requires sophisticated sensors and software to work on the go. Calibrating fixed CNC machines is considerably cheaper as the machine and workspace stay fixed in place, and the workpiece - the growing pods - are brought into the workspace. A basic CNC machine 500 has a fixed frame, motorized movement systems for the X, Y, and Z-axis and the end piece which performs the task. Wood manufacturers, for example, use CNC machines with a rotating mill installed at the end piece. The wood is placed and calibrated to the workspace of the CNC machines, which executes a recorded set of commands to mill the wood to specifications. When performing robotic precision farming, the wood is replaced with growing pods and the mill with a seeder/watering/replanting (etc.) head. While most of the precision farming tasks can be executed with a 3-axis CNC machine (that can move within the X, Y, Z-axis of a fixed space), some tasks, such as pruning, require more advanced 5-axis CNC machines. These machines have additional motors that can move the end-piece (or the workspace itself, which is not suitable for growing pods full of liquids) so that it can reach additional locations. The additional motors allow the changing of the angle of attack (up, down) of the end-piece and rotate it around. In many ways, a 5-axis CNC machine is similar to robotic arms, which allow wide freedom of movement. For growing pods, which are always tasked on the same horizontal axis, a 5-axis CNC machines may be used for complex tasks rather than robotic arms. This is due to the additional costs for robotic arms because of advanced calibration sensors required and the far more complex design. The end piece of the CNC machines 500 can be fixed, supporting the specific precision farming tasks the piece was designed for. For example, the pod cleaning CNC machine would have a fixed head with rotating brushes and water/detergent sprayers. Alternatively, the CNC machine can support interchangeable end-pieces, allowing the machine to perform a variety of farming tasks, as well as easily upgrade and add parts.

[0095] With reference to Fig. 16, a CNC machine 500 with a fixed cleaning head 520 (due to water pressure, temperature, and/or the use of detergents, which are not food safe) may be used for cleaning. The cleaning head 520 may use a combination of high- pressure water, cleaning detergent, and rotating brushes to clean, for example, the pods’ root chamber, nutrient reservoir, and the various lid tops. With reference to Figs. 17A and 17B, in some embodiments, one or both of the lids 12 and the containers 10 of the pods 100 may be designed with hooks 110, 120 to allow the CNC machine 500 to pick up the pods 100. This allows the CNC machine 500, together with the rovers 200, to prepare and place the containers 10 and lids 12 for new growing cycles.

[0096] When using soil-based pods, new or cleaned containers 10 may need to be filled with soil ahead of seeding. As illustrated in Fig. 18A (Add Fig. 18 A), a CNC machine 500 may use a pipe-like head 530 A to add soil into the container 10. The same head 530A may also be used to suck the used dirt from containers 10, which may then go through a cleansing process before being recycled and used again. When using hydroponic or aeroponic pods, the seeds may be placed into grow cups 80A or 80B (see Figs. 9A, 9B) which include a medium (such as coconut, Rockwell or fertile). As illustrated in Fig. 18B, a pipe-like head 530B (similar to head 530A used to add soil) may be used to fill the grow cups with the medium. In hydroponic and aeroponic pods, the lids 12 are used to hold the grow cups above the container/root chamber. Soil pods use lids 12 for drip irrigation to each of the growing crops to minimize water waste (see, for example, Fig. 7B). Different lid layouts may be required to maximize the farm’s density. The system may change the lids 12 during the growing cycle in different stages. These lids 12 may have hooks 120 (see Fig. 17B) for the CNC machines 500 to connect to and lift. CNC machines may be customized to perform seeding. Figs. 18C and 18D illustrate different embodiments of CNC machines 500 that may be used to perform seeding. In the embodiment of Fig. 18C, a seeding head 540 A pushes one seed through a tube into the grow cup. In the embodiment of Fig. 18D, a seeding head 540B seeds an entire row of grow cups simultaneously. In a preferred embodiment, the seeding head may use a suction system with multiple needles to grab one seed and place it within the growing medium. Suction heads may be suitable for pelleted seeds. For farms using non-pelleted seeds, vibration heads may be used. The vibration heads may vibrate the seeds so that only one seed drops down the placement tube. Fig. 19C illustrates a CNC machine 500 with a vibrating platform 510 that the seeds are picked up from. The seeds are vibrated so the CNC machine 500 can easily pick them up. [0097] CNC machines 500 may include cameras for both calibrating the robots and crop analysis and monitoring. Calibration cameras may be placed around the frame, while the crop cameras (and precision calibration cameras) 550 (see Fig. 18D) may be placed near the end-piece of the CNC machine 500. Different filters and lenses installed on the cameras may serve as spectrometers, which allow the software to monitor the crops’ growth progress, damaged leaves, and potential diseases. Root health is critical for optimal yields and growth rates. By combining the replanting end-piece (which allows the CNC machine to pick up and move crops between crops) and the cameras described above, the system can pick up the crops and photograph their roots. Root- related tasks may be performed at night as roots, when exposed to sunlight, go into shock and slow the growth of the crop. Non-photosynthesis light spectrums may be used to assist robotic navigation. These spectrums do not harm the roots as they do not provoke chemical reactions.

[0098] To maximize farm yield, crops are replanted throughout their grow cycle. Replanting allows the precise allocation of minimal space and maximum density, with more space made available as the crop grows. As illustrated in Fig. 18E, replanting may be performed by hooks 560 of the CNC machine head that engages with grow cups 80A. As illustrated in Fig. 18F, an electromagnetic head 560B of a CNC machine 500 may also be used for replanting a grow cup 80B. Replanting is also useful for robotic attending of defective pods, for example, when their pumps stop working. The robotic storage system can bring the problematic pod to a replanting CNC machine and fetch an additional empty pod. The CNC machine will then replant the crops to the new pod with the working pumps. This allows the autonomous operation of the farm with only minimal urgent human maintenance necessary. Sick crops, for example, can quickly be identified by the camera heads and software, and then, similar to replanting those crops, can be picked up and discarded.

[0099] Seeds often require water to jump-start their growth. In some embodiments, a watering CNC head 570 may be used to deliver water or a nutrient-rich mixture to the crops. In some embodiments, the head 570 may deliver water to an entire row of crops simultaneously. In some embodiments, the watering head may deliver water to some or all the crops in a pod simultaneously. Watering is especially critical in hydroponic and aeroponic pods, as the nutrient delivery system is placed within the root chamber, and initially, the crop does not have hanging roots with access to the nutrient solution. Watering is also useful for soil pods to ensure the seeds are placed in a humid, rich soil so they will successfully germinate. The watering system may use the same nutrient management system explained later. When using pelleted seeds, water alone can be sufficient for germination, as the coating includes the initial nutrient compounds required for the seed to germinate.

[00100] When using soil pods, in some embodiments, sensors are installed as an end- piece of the CNC machine 500. The end-piece can probe the soil. Fig. 18H illustrates an exemplary embodiment of CNC machine 500 with a plurality of sensors 580A, 580B, etc. attached as an end-piece on the head of the CNC machine 500. These sensors may include one or both of humidity and conductivity sensors. These sensors may acquire signal indicative of the health of the soil, for example, signals that indicate whether the soil has sufficient moisture from the nutrient solution. In some embodiments, additional compound-specific sensors may, alternatively or additionally, be placed as an end-piece to identify the exact combination of nutrients currently available in the soil.

[00101] The disclosed system may be configured for testing, adjusting and managing the nutrient reservoir of the pods 100. In some embodiments, managing the nutrient solution may be performed using a separate device explained later. Tubes may be placed on an end-piece of the CNC machine 500 and connected to the nutrient system with pumps.

Fig. 181 is an illustration of two tubers or pipes 610A, 610B being lowered into the nutrient reservoir 60 of a pod 100 by the CNC machine 500. The pipes 610A, 610B are used to transfer nutrient solution to and from the pod 100 to the nutrient management device. The CNC machine 500 may place the pipes 610A, 610B precisely within the pod’s nutrient reservoir (the location of the reservoir changes depending on the growing method used). One tube may be used as the “dirty” tube, which can pump the solution from the pod reservoir to the nutrient system. This is done to sample and test the nutrient solution or to completely drain the reservoir when required. A “clean” tube line is used to pump the solution from the nutrient system to the pod reservoir. This line is used to pump an adjusting nutrient mix or refill the pod when needed. The tubes are separated into a dirty and clean line to minimize any cross-contamination between two different pods 100, which are serviced one after the other. If contamination occurs in one pod 100 and then another containment-free pod 100 enters the CNC machine for maintenance, the contaminated solution will not be pumped into the pod, but only to the dirty chamber of the nutrient system. To clean the “dirty” lines, an empty pod/reservoir is used. Water and detergent are pumped via the “clean” tubes, from the nutrient system to the pod 100, and then back to the nutrient system through the dirty sampling “dirty” line. After a few cycles, both tubes are cleansed and contaminant-free. Alternatively, nutrient sensors (such as PH, EC, and oxygen) may be mounted on a CNC end-piece (see Fig. 18H) and lowered into the nutrient reservoir, thus allowing the sampling of nutrients without the separate nutrient system.

[00102] A CNC machine 500 may also be used for pruning the crops in pods 100. Pruning is mainly required for larger crops, such as tomato crops, where leaves and parts of the stem are removed to direct the crop’s growth direction or to maximize the quality and yield of each crop. As illustrated in Fig. 18J, automated pruning may be performed using a scissor-like CNC end-piece or pruning tool 620. An air vacuum tool 630 (see Fig. 18K) may be used along with the pruning tool 620 to remove the trimmings. Similar to industrial CNC machines, fully automated pruning may require the pruning tool 620 to have freedom of movement on other axes (beyond the X/Y/Z-axis). This may be done with additional motors installed on the end-piece, which allow the piece to move around in place and change its angle of attack.

[00103] A CNC machine 500 may also be used for trellising the crops in pods 100. Fig. 18K illustrates a trellising tool 640 used for trellising crops. Trellising, required mainly for larger crops such as tomato plants, is performed by placing support structures around the growing crop to guide its growth and provide support for the increasing weight of the flowering crop. This can be performed via replanting the crop to a custom grow cup with support structures. Similar to pruning, fully autonomous trellising may require advanced CNC end-pieces, such as trellising tool 640, with 5-axis movements to access every part of the growing crop.

[00104] A CNC machine 500 may also be used for trellising the crops in pods 100. The type of harvesting performed may depend on the type of crops grown. Fig. 18L illustrates an exemplary harvesting tool 650 being used to harvest crops. In some embodiments, multiple harvesting tools 650 may be attached to the head of the CNC machine 500 to simultaneously harvest multiple (e.g., all) crops in the pod 100. Harvesting tool 650 may resemble mechanical scissors that cut the crops from its roots, while a mechanical grabber (positioned above the scissors grabs and holds the harvested crops. For microgreens, the harvesting tool 650 may include a CNC end-piece with a blade - that is placed above the lid of the pod - is moved to cut the microgreens from the roots. The microgreens can then be collected via a vacuum pump or with a conveyor belt, which collects the harvested microgreens. For leafy greens, such as lettuce and basil, an end-piece that both cuts the bottom part of the crop (to separate it from their roots) and then holds the cut crop, which is placed aside for packaging. For fruits that require precision harvesting, a 5-axis CNC end-piece is required.

[00105] The harvested crop may be placed into a bag and then sealed. In some embodiments, a CNC end-piece may be used to open the packages and place them in a specific “packing” pod. The pod is then brought to the harvesting station, which grabs the crop and places it in the open and waiting package. The pod is then returned to a CNC machine with a packing tool which seals the bag with heat.

[00106] In some embodiments, to support so many different tasks, without installing a specific machine for each task, an interchangeable CNC end-piece may be used. This system allows the CNC machine 500 to interchangeably use multiple heads. That is, the current head (e.g., with a pruning tool 620) may be unmounted and a different head (e.g., with a harvesting tool 650) attached and locked into the CNC machine 500. An interchangeable locking mechanism may be required to secure the end-pieces to the CNC machine 500. The interchangeable locking mechanism may include magnets, a screw system, or a vacuum system, which securely holds the different end pieces to the CNC machine 500.

[00107] The disclosed system may include a nutrient management system 700 (see Fig. 20) that is configured to perform nutrient management. The basic compounds of the nutrient solution may include nitrogen, phosphorus, potassium, calcium, magnesium, sulfur, iron, boron, magnesium, copper-zinc, molybdenum, chlorine, and nickel. In soilless methods - such as hydroponics and aeroponics - all nutrients are supplied to the plants by the nutrient solution, a well-prepared liquid fertilizer. The following conditions may be preferred for an optimal nutrient management system: all elements must be contained in the solution; all elements exist in ion form; the concentration of the different elements is within the appropriate range for normal plant growth; solution is clean from harmful substances and pathogenic microorganisms; the pH is stable around 5.5-6.5; enough oxygen is dissolved for root respiration. The nutrient management system may be comprised of the following main components: purified water tank + reverse osmosis filtration system; nutrient tanks + precision pumps that feed the solution to the clean mixing tank; a mixing tank where the water and nutrients are mixed together, oxygenated and then tested with compound-specific sensors to verify the recipe; a sampling tank where the nutrient solution from existing pods is pumped into and tested with compound- specific sensors to ascertain the status of the solution; and pumps to pump from/to the nutrient system to the nutrient reservoir of the pods through tubes that are installed on the end-piece of a CNC machine.

[00108] The basic solution begins with purified water, which goes through reverse osmosis. This removes any existing compounds that may affect the crop’s normal growth. Additionally, heavy substances that exist in drinking water may be removed to minimize the damage to the nutrient distribution systems in the pod (such as ionizers, pumps, or spray nozzles). When water is recycled between pods, the solution is purified again, with the addition of UV lights that kill any bacteria (which would not exist in the drinking water due to the chlorine, and which has already been removed in the first reverse osmosis process). The purified water is stored in a separate tank available for the nutrient system (a separate tank is required as the purifying process is slow, so better to allow it to run constantly and have a large amount of purified water available when needed). The above-described nutrients may be supplied as separate compounds or using pre-mixed solutions, which have some of the required compounds at their proper ratios. Using pre-mixed solutions minimizes the number of pumps and sensors required for the nutrient management system. The nutrient tanks may be filled from time to time by the farmer, and level sensors - installed within each of the nutrient tanks - may trigger a notification, which will be sent to the farmer when a refill is required. The nutrient management system may prepare the nutrient solution by pumping purified water into the mixing tank and activating precision pumps (which allow pumping small and precise amounts) to add the different compounds or nutrient-mixes. The solution is then thoroughly mixed via an air pump or a magnetic stir system. Sensors, such as pH, EC, and oxygenation may be installed within the mixing chamber to verify that the prepared solution has the correct readings as per the grow protocol. Once the solution is verified as correct, it is pumped from the nutrient management system to the pod’s nutrient reservoir.

[00109] A different protocol can be used for different crops and pods. To sample the solution within the nutrient reservoir of a pod, a small quantity of the solution is pumped out of the reservoir and into the sampling chamber of the nutrient management system 700. Sensors - such as pH, EC, oxygenation, and compound-specific sensors - analyze the current solution. The system, together with the software, is able to prepare a custom nutrient solution mix, which, when added to the nutrient reservoir, will correct the existing solution to its correct values. For example, if the acidity is high, a base compound and water can be pumped into the reservoir and mixed to adjust the solution to the correct acidity levels of 5.5-6.5. Once the adjustment solution fills the reservoir, an additional sample is sent back to the nutrient system to verify the mixture. If the system is unable to adjust the existing solution in the nutrient chamber, the entire chamber may be drained (and then recycled back to purified water for other pods) and refilled with a new mixture.

[00110] The cloud software (in control system 400) receives the sensor readings from the nutrient chambers and sends the commands to activate the pumps (either to the pod or from the nutrient compounds) and to mix a tailored and specific nutrient solution.

Another use of such a flexible nutrient system is that custom solution recipes can be used to compensate for non-optimal farm climates. Additional compounds can be given, which help the crop grow even when the climate is relatively warm. The software can calculate and choose between cooling the farm via climate control systems, such as air conditioning, or instead, changing the nutrient solution recipe to better cope with the existing climate. [00111] The nutrient management system 700 may also be used to develop new nutrient solution recipes and growing protocols. The A.I. powered software can self-experiment, change the solution, and then monitor how the new recipe affects the growth of the crop. The nutrient solution is only part of the testing process for new protocols. The environmental conditions are also changed, in parallel with the nutrient solution, to find an optimal combination of all factors.

[00112] To use and optimize the robotic growing system efficiently, the system may include software configured to control the growing system. This software may be split into the following components: embedded IoT software, robotics software, cloud software, and a self-teaching artificial intelligence (AI) software. The embedded IoT software may be installed on the controllers (e.g., microcontrollers 26) of the growing pods 100. These controllers broadcast data from sensors in the pods 100 to the cloud software and receive commands back. Lights 42 (see Figs. 10A, 10B), either installed on each pod 100 or as a separate spacer, may be controlled with the same software. The robots of the discloses system include the mobile rover 200, forklift 300, CNC machines 500, and nutrient systems. These robots include software. The moving robots in the growing system require more computational power and advanced software for complex tasks. These robots may use cameras and/or radar sensors to calibrate their location visually. To perform the synchronized movement of multiple robots (for example, to unpack a stack of pods), the software must support a large buffer of present commands, and fast local connectivity and processing of data between the robots.

[00113] The control system 400 may include cloud software. The sensors in the system (in pods 100, rovers 200, forklift 300, CNC machine 500, etc.) may broadcast their data to control system 400 which may be a remote cloud management system. This system may also create dynamic commands to the farm systems, based on internal and external data, such as market demand or weather forecast. A self-teaching A.I. may be used to optimize the growing protocols of various crops in various conditions.

[00114] The pod storage software, also known as AS/R — Automated Storage and Retrieval — may be used to manage the robotic storage and stacking of pods in the climate-controlled growing area. The AS/R software may be installed locally or in some embodiments as a remote service where the software runs in the cloud (remote servers hosted away from the farm). The software manages the locations of each pod (on the floor and within each stack) and creates the optimized command sequences to store or retrieve pods. These sequences of commands may use one or more storage rovers and forklifts. Stacked pod storage may be similar to “puzzle-based systems,” in which there are many ways to store or retrieve a pod, and the most optimized way needs to be calculated and found. The software may, in some embodiments, simulate and identify the bottlenecks of the robots (such as the time needed to move stacks out of the way before the wanted stack is reachable) and use those calculations to coordinate the other tasks before and after such bottlenecks. Such an embodiment will allow a more constant flow of robotic operations with fewer robots needed per farm.

[00115] The software may manage the floor space and the possible locations to park stacks of pods 100. The storage plan is flexible and can change dynamically.

Limitations, such as height in specific spots, may additionally be inputted to the software. The software also supports multiple growing spaces that are connected or accessible to move between through robotic rovers. This, for example, is used when most of the growing space is indoors, but a portion of naturally lit space is also made available. The cloud software can manage the movements of the pods between the areas, and also optimize which pods will receive the natural light (based on their crop and stage). This decision may consider different parameters depending on the preferences of the farm owner, such as to prioritize yield or minimize energy consumption. A weighted decision based on both is also possible.

[00116] In a stacked farm layout, pathways to the stacks of pods 100 are not required. Instead, stacks can be hidden between other stacks (see, for example, Fig. 11). By moving the stacks, the hidden stacks may be made available. In theory, the farm may be almost completely filled with pods without any pathways. In this type of layout, retrieving a stack from the back of the farm will require a lengthy and costly sequence of rover movements. To retrieve a stack that is hidden by many other stacks, rovers 200 move the stacks and slowly brings a desired pod 100 to wherever it is required (e.g., to a CNC machine with a harvesting tool 650). The cloud software may use the farm data, as well as external data, to dynamically build the most efficient storage plan. For example, seedlings often require little to no maintenance for the first few days of growth. This means a stack of pods with fresh seedlings is an optimized candidate to hide between other stacks. Parking stacks, which will require maintenance at similar times, near each other, will further reduce the overall needed rover movements and save on energy consumption.

[00117] The AS/R cloud software may also shuffle pods between different locations on the farm or within a stack. For example, if the growing space is a controlled climate greenhouse with natural sunlight. Sunlight is the most efficient lighting source for photosynthesis, but is primarily effective for the top pod on each stack, as that same pod shades the pods beneath it. To optimize the use of light, the AS/R software may send commands to the storage rovers and forklifts to change the order of the pods on the farm. In addition to the optimization of light, different storage layouts and heights may be used, which affect the climate and airflow within the farm. The AS/R system, in some embodiments, may use external data — such as the weather forecast — to dynamically adjust the storage layout.

[00118] The AS/R system may be used to work in conjunction with human employees. Retrieval tasks for humans can be prepared around their working hours, as the storage system may autonomously operate 24/7. Blocking prevention is another aspect that the software can solve. Multiple retrieval/storage paths may exist in case of broken-down robotics in the growing area, or if other blacking tasks are currently being performed nearby. Path planning may be optimized, as well as the parking locations of the storage rovers, based on the assumed future tasks of each robot. The AS/R software may keep track of the current and historical location of each pod and crop within the pods. This is beneficial for situations when a pest is found on a crop. In such cases, nearby crops can be retrieved for monitoring and disposal if required, before infecting other crops in the farm.

[00119] The cloud software may incorporate a large number of features that optimize the yield and revenue of the farm. These features are described below.

[00120] Replanting: crops may be robotically replanted between pods. For example, seeds are placed at 100 seeds per 1 SQM and later replanted — as they grow and require more space - to pods with lower densities per SQM. This allows each crop to be dynamically allocated in the minimal space required through its growth cycle.

Replanting requires a coordinated sequence of two or more pods and one or more storage rovers, forklift, and a replanting CNC machine.

[00121] Future Simulation: the software optimizes the farm density based on the future space required as the crops grow. This feature allows for the intelligent coordination of seeding new pods to keep the farm, over time, at the highest possible density of crops.

[00122] External data: the software uses a tremendous amount of external data to individually optimize each farm. This data includes weather forecasts that are used to prepare for climate changes, like shuffling the location of pods, changing their nutrient mixtures, or changing the configuration of the climate control systems. E-commerce data is used to recommend the most profitable seeding plans based on consumer demand and preference (such as holidays, vacations etc.). News of supply shortages can be additionally used to change the farm plan.

[00123] Price awareness: the software uses external data, such as energy, water, nutrient, and labor costs for price awareness. This data, combined with the projected revenue models, allows the software to optimize the profitability for each farmer and farm. For example, different artificial lighting schedules can be tailored for different locations, based on the local energy costs throughout the hours of the day.

[00124] More sellable parts: the system optimizes the growing protocols, pruning, and exact harvest time to maximize the overall sellable part of each crop. This is achieved based on historical data and testing, as well as via machine vision comparison.

[00125] More favorable crops: having almost full control over the climate, nutrient composition and lighting regime of the crops, the system can grow consistent and replicable crops, over time, and in different locations and climates. Medicinal plants, for example, are highly valued if the crop is replicable over time. This allows the medicine manufacturer to better forecast their demand for the plants. Consumers, for example, often favor specific shapes, colors or flavor profiles for certain crops. This preference often changes between communities and locations. The software is able to optimize the climate and recipe to produce crops that are the most favorable, thus generating a price premium and additional profit for the farmer.

[00126] Faster, larger crops: the software uses both research growing protocols and self- optimized AI grow recipes that change the growing conditions of each crop. This includes setting the pod type, nutrient composition, dosing and lighting regime, and the sourcing climate to grow the most cost-effective crops. Specific nutrients, for example, can be added in specific stages of the crop’s cycle to boost its growth rate and mass.

[00127] Consumer feedback: as each crop grown has a unique ID and an entire log of the conditions, protocols, and images used during its grow cycle, consumer feedback can be used after the crop is harvested and sold to optimize crops to the consumer preference. A QR code that forwards the user to a crop-specific survey can be added to the packing of each crop to gather this information. This information allows for further customization and optimization for the desired local taste profile.

[00128] Recycling resources: the software tests the used nutrient solution from harvested pods. Based on that data, the solution can be filtered back to clean water, used as-is for a new crop cycle, adjusted with new nutrients, or disposed of. The software is cost-aware and takes into account the energy, water, and nutrient costs.

[00129] Maximum storage density: as explained in the previous AS/R chapter, the software uses internal and external data to optimize the dynamic storage plan of pods on each farm.

[00130] Minimize crop failures: the software can identify broken-down pods based on images and sensory data. When growing in aeroponic pods, for example, if the misting system stops working, the roots will begin to irreversibly dry within 2-3 hours. Even if no humans are present on the farm, the system is able to unstack the problematic pod, replant its crop to a new working pod, and move the pod aside for future human maintenance. A custom nutrient protocol can be used to help revive the dry roots.

[00131] The cloud software incorporates a large number of features that optimize the energy use of the farm. These features are described below. [00132] Circulating pods: as natural sunlight is mainly accessible by the pods stacked highest on each stack, the software can circulate and shuffle the pods to best distribute the sunlight based on the current crop and its stage. A similar optimization can be done with the artificial lighting installed throughout the farm, for example, calculating and taking into account the ambient light exposure from the different light sources. The same method is also useful when shading is required, for example, for seeds or medicinal crops, which are initially put into a shock (dark) state to optimize specific reactions.

[00133] Efficient climate management: different crops prefer different climates. This is why crops naturally do not grow everywhere or year-round, as the conditions are only sometimes suitable. When growing in a controlled climate space, this can be circumvented by changing the indoor climate. However, another alternative is changing the nutrient composition of crops. For example, specific compounds can be added when the climate is too warm to help it grow nominally even if the climate is not. The software weighs the costs and advantages between changing the indoor climate or the nutrient composition and finds a cost-aware and optimal solution.

[00134] Timing tasks: robotic tasks can be prioritized for the low-demand hours when energy is often considerably cheaper. This dynamic schedule can affect both the lighting regime in the farm and when the robotic tasks are executed.

[00135] Alternate climate systems: based on the energy cost and the required climate change, an optimized solution for the use of different climate systems - such as heat pumps, radiators, air condition, and humidifiers - can be found and used.

[00136] Reduce robotic movements: the timing and sequencing of the various robotic tasks can be optimized to use the least amount of electricity.

[00137] Optimal storage plan: while the pods can be stacked in extreme densities, the optimal plan will often be considerably more spaced out. For example, specific plans can achieve better airflow in the farm, thus reducing the required cooling from expensive systems. Ambient sunlight can be better distributed at lower storage densities, thus minimizing the need for artificial lighting. [00138] The robotic storage system is designed to minimize the required human labor per kilogram of crops considerably. This is achieved by the AS/R storage system as well as other software features described below.

[00139] Humans in place: the human employees stay in place, and the storage rovers retrieve and bring the pods to the humans for maintenance. This is also done for the robotic precision farming tasks, and the reason why we are able to use stationary CNC robots. This reduces the wasted travel time of the employees, and maximizes their productivity. As humans do not normally enter the growing space the climate conditions can be optimal for crops with high levels of CO2 (which are unhealthy for humans).

[00140] Plan human tasks: in addition to directing the robotic systems and devices, the cloud software also plans the human task list. The human task list is synced with the robotic system, thus allowing an ongoing flow of tasks with little to no bottlenecks. Cameras and image recognition can also be used to visually aid the human employees on how to perform the task at hand. For example, when pruning, the software can produce a visual overlay that shows where the incisions on the crop should be made. This minimizes the required farming expertise of the human employee while maintaining the high level of successful crops. Human tasks can also be planned, so they are required only during business hours. The expertise level of the human farmer can also be taken into account, and propose a suitable seeding plan with suitable crops and tasks.

[00141] Robotic tasks: a large and growing number of farming tasks, traditionally required to be performed by human farmers, are executed via autonomous robots. The storage rovers and forklift allow high planting density and optimal nutrient absorption provided by the IoT pods. With the addition of the stationary CNC robots, delicate tasks such as nutrient management, seeding, cleaning, pruning, and harvesting can be robotically automated. This will leave the human employees mainly responsible for robotic troubleshooting and maintenance.

[00142] Image recognition: images and machine vision is used to verify the successful execution of robotic tasks. For example, when a CNC robot seeds a pod, cameras and images are used to verify the successful catching of a seed and then the successful placement of the seed into the growing medium. This allows close to 100% successful seeding and minimizes unused growing space on the farm. Images can further be used for the software to self-monitor the crops as they grow. Patterns identified can be used to adjust the growing protocol or create an alert if a significant variation is detected.

[00143] Human UI: the software has an extensive user Interface, as the farm is designed for parallel work between humans and robots. This management system is used both to set the seeding plan, the climate condition, and input new data. The system then provides ongoing monitoring and control of the farm to the user, as well as provides a recommended human task list and the detailed instructions to perform them successfully.

[00144] Remote monitoring: the entire system can be remotely monitored either by the farm owner or by a central service provider. As the farm and its systems are completely integrated with the cloud software, remote 24/7 human monitoring is possible.

[00145] To minimize the hardware costs of the system, simple and affordable electronics are used. Limited microcontrollers, such as STM32, ESP8266, and ESP32, may be used due to their low cost. These devices are suitable for the pods, artificial lights, and forklifts - systems that are not speed or latency sensitive due to the low number of commands they perform. Storage rovers and CNC robotics require different, more capable software and hardware.

[00146] The embedded software installed on these microcontrollers provides the ability to remotely flash the microcontroller with new firmware, and connect to the wireless network on the farm. This can be a WiFi, LORA, or custom RF network. Alternatively, a wired connection can be implemented with the addition of a wired network modem module. The software can further allow a local mesh network between the devices (thus minimizing the required central network hardware requirements). The embedded software also enables or provides the following features.

[00147] Interpret the sensor information. Drivers are implemented to support the various sensors, such as weight, water level, and humidity above the pod and within the root chamber, and the LUX, UVA, UVB, UVI, and spectrum analysis of the light reaching the pod. The sensors vary based on the type pod. For example, soil pods require a soil humidity sensor instead of an air humidity sensor, which is required for aeroponic pods. [00148] Broadcast and receive commands from the management system. A preferred embodiment is of a cloud MQTT service that can support a tremendous scale of devices and farms.

[00149] Fallbacks when the network is not available. This can include internal sequences, such as pumps or lighting schedules, so the crops continue their growth during blackouts.

[00150] Energy monitoring and management is used to measure and optimize the energy consumption of each pod and device. This can also be used to remotely identify damaged devices. The software also supports deep-sleep, where energy consumption is minimized while the device is idling.

[00151] Control the pumps or ionizers for the nutrient dosing system or the motors and feedback systems of the forklifts or storage rovers.

[00152] New farm setup: the device firmware is designed to be quickly identified, tagged, and updated when a new farm is installed. This is achieved by the devices initially broadcasting their own network where they are waiting for commands and a unique ID allocation.

[00153] Triangulate location: based on the signal strength from stationary beacons or other devices.

[00154] Run dynamic rules: for example, the pod IoT microcontroller can monitor the humidity level within the root chamber and activate the dosing pumps whenever the level falls beneath a specified threshold.

[00155] Stationary CNC machines are a preferred embodiment for performing the precision farming tasks required by the crops and pods. These machines are remotely controlled by a microcontroller that connects to the local wireless network and the farm, and through that network to the central cloud management software, which directs the various tasks. The CNC microcontroller supports the various motors and feedback sensors that move and operate the CNC end-piece. Additional sensors, such as proximity sensors, can be used to calibrate the robot with its workspace and parts (e.g., the growing pods). An additional verification and calibration process may be performed when the end-piece is changed.

[00156] Fixed cameras on the CNC frame and head are used, in combination with machine vision software, to help calibrate the pods 100 and crops. Once the images are processed, a sequence of movements - tailored for the exact location of each crop and leaf - is sent for the CNC machine to execute. Additional images may be used to verify the successful execution of specific tasks, such as seeding, replanting, and cleaning. Some CNC heads may have additional sensors that may be supported by the CNC microcontroller. These sensors can include nutrient reservoir sensors (PH, EC, oxygen, temperature, etc.) or soil sensors.

[00157] As the CNC machines 500 rely fully on the storage rovers 200 (and forklifts 300) to retrieve and bring the pods 100 which require maintenance, the cloud AS/R software is synced with the planned CNC tasks, so the pods 100 are brought within the correct time slots to allow a continuous robotic maintenance flow. The CNC machines 500 may also be used to take close-up photos of the crops within each pod 100. The images are then processed for analysis and monitoring, as well as a historical data-set for AI training. The embedded CNC software also supports the storage and execution of sequences as tasks. This allows the cloud software to send larger amounts of data and allow faster execution, without network latency between each task. The software supports different machine types, motor drivers, and feedback sensors.

[00158] The growing system makes use of automated nutrient management systems and machines. These machines can precisely mix a custom nutrient solution from a large number of base compounds. Sensors installed in the mixing tanks are used to sample and verify the solution composition. Sensors can be used to identify small traces of the different compounds or to measure electric conductivity, acidity, oxygen saturation, or the temperature of the solution. These sensors, in combination with the cloud software, are also used to adjust the existing nutrient solution within the nutrient reservoir of each pod. This feature reduces the resource and water waste, as the solution is adjusted instead of being drained and refilled like existing commercial solutions. The cloud software first identifies which pod is being serviced, and then adjusts the nutrient settings based on the currently used growing protocol. If the nutrient solution is considerably out of range, the reservoir will be drained and refilled with a new solution, as adjusting the existing solution may require more resources and nutrients.

[00159] Growing protocols are instructions for growing different crops. Different protocols exist for different optimization goals. For example, energy-optimized protocols use less energy but require higher nutrient use. Revenue optimized protocols will grow and replicate consumer-favored crops, but energy use will be higher due to the stricter control required of the climate within the farm. These protocols, which change for each crop and through the life cycle of the crop, set the following parameters:

[00160] Nutrient solution composition - how much of each compound is used and what levels must be maintained in the reservoir. How often does the nutrient solution need to be sampled?

[00161] Nutrient dosing - the regime of activating the dosing pumps or ionizers. This regime can be dynamic and take into account sensory data such as humidity levels within the pod.

[00162] The lighting regime: how much light, and which spectrums, are needed for optimal crop growth. This also affects the use of natural/artificial lights and their timing and regime. The dynamic protocol can take into account weather conditions, forecast, and energy costs.

[00163] Replanting: when more space is needed and the crops need to be replanted into lower-density pods.

[00164] Pruning: what crop shape is recommended, and whether it should be pruned.

[00165] Climate settings - such as temperature, humidity, C02 and airflow that need to exist in the growing area for optimal growth.

[00166] Problem response: the protocol also recommends different responses based on problems in the optimal conditions (for example, machine vision recognition of different shape/color, broken dosing system, or unforeseen low sunlight levels). [00167] Climate specific: different protocols exist for different climates. For example, warmer farm locations will use a different nutrient solution that better combats non- optimal heat conditions.

[00168] Growing protocols allow replicable and consistent crops across different farms and locations. This is especially important for medicinal and premium crops.

[00169] The disclosed robotic growing system may be configured for artificial intelligence. The adaptive A.I. generates dynamic growing protocols and commands the robots on the farm. The adaptive commands can optimize the yield, revenue, or resource use of the crops and systems. Initially, scientifically researched growing protocols for each crop are inputted into the system. These protocols recommend bands for most of the controllable variables of the pods and crops, as explained previously.

[00170] To optimize the base protocols, R&D pods are used. These pods allow farm- level climate control, but for individual pods. This allows not only the changing of the nutrient, dosing, and lighting protocols (as is possible for all growing pods), but also the ambient climate temperature, humidity, C02, and airflow. These R&D pods also allow the AI software to simulate the climate in different geographical locations of future farms. To optimize the growing protocols, the AI software creates different tests with different protocols to find the most efficient, replicable, and effective protocols. The tests are initially done within the scientifically recommended bands, but also outside of the bands to test varying combinations of controls. The software is also able to self-teach itself the most successful protocols, based on sensory inputs like weight and composition and through machine vision from images. Human feedback, provided for the different crops, can be further used to optimize consumer-favored crop protocols.

[00171] Unlike conventional rule-based programming, the AI software uses an evolving neural network that generates different commands based on the changing inputs. The exact decision process of the software may be unclear to humans, but it is based on the data-set of historical tests and experiments. The AI model can also be used to approximate the consumer demand based on external data, and formulate effective grow plans for each farm, based on all input data. [00172] The data from the different farms (e.g., in different locations) is also shared and used to train and advance the AI. This allows the AI to recommend and execute different commands based on data seen elsewhere and a dynamic best practice. The software can run locally or in the cloud. Running locally is especially useful for machine vision and image recognition, due to the high bandwidth required to broadcast to a remote server.

The AI model can be easily updated remotely as it advances. The evolving growing protocols, seeding, and storage plans allow the software to optimize farms over time and provide additional yields and revenues with software changes alone.

[00173] Image recombination and machine vision are used throughout the robotic growing system. The software that is used to process the images is recommended to run locally on a server in the farm, as fast internet connectivity is usually less available in farming locations. Additionally, as many robotic sequences require multiple machine vision verifications during the sequence, running the software locally will lower the latency and speed of the robotic tasks.

[00174] Cameras are installed on the storage rovers 200, forklifts 300, CNC machines 500, and in fixed locations throughout the farm. The images taken from these cameras are used for: (a) locating other robots and devices; (b) allow the autonomous navigation for the storage rovers and movements of the forklifts; (c) create a precise 3D point map for the CNC machines to perform precision robotic tasks; (d) monitor the pods and crops. Images can be compared to historical data-sets and baseline images to identify and alert if the growth is different; (e) Identify fungi and diseases. This also allows the robotic disposal of infected crops, which lowers the chances of cross-contamination; (f) verify robotic tasks: such as seeding, in which an image is processed to verify a seed has been placed in the correct location and at the right depth; (g) verify replanting, and that the crops have been successfully moved from one pod to another; (h) optimizing harvest time: when the sellable parts are maximized and when the shape of the crop is most favorable for consumers; (i) replicable crops: identify and compare crops to deliver replicable crops.

[00175] To achieve optimal crop growth in different climates and locations, the disclosed growing system includes automated climate control in the growing space. The dynamic growing protocols may, in some embodiments, also actively change and control the climate conditions based on the crops, and their stage/condition, growing on the farm. In alternative embodiments, the dynamic growing protocols may change the nutrient composition within the pods to better cope with the external climate conditions. In a preferred embodiment, the dynamic growing protocols weigh both options (changing nutrient composition or activating climate control) based on external cost data of the nutrients and energy and make an optimal decision. As different crops have different optimal climate preferences, the A. I. powered software may make smart recommendations for which crops to grow, so the overall required climate is similar throughout the farm. Alternatively, the system may recommend specific crops that will be the most energy-efficient to grow, based on external data such as the upcoming weather forecast.

[00176] The stackable growing system is suitable for controlled climate farms, such as indoor, greenhouses, and glasshouses. A controlled climate farm may require an isolating layer that keeps the internal temperature and climate conditions from leaking outside. Additionally, electric systems, such as air conditions and C02 dispensers, are required to reach the optimal conditions.

[00177] Soil-based pods require the least amount of climate control and may be suitable for stackable outdoor farms without any isolation of climate systems. Indoor farms are fully enclosed and isolated from the outside. They may have, in some embodiments, windows that allow the natural light to enter while keeping the external air from entering. These farms may be located in warehouses, depots, and parking lots due to the software’s ability to support dynamic floor plans with little restraints. Due to the limited or non existent sunlight, indoor farms rely heavily on artificial lighting systems (as described earlier, either under the pods or via lighting spacers). By having full control of the lighting regime, dynamic protocols can emulate different climates or sunrise/sunset times.

[00178] An advantage of indoor farms is that they offer excellent climate isolation, which lowers the overall energy consumption of the climate control systems. In some embodiments, the internal air may be filtered, which lowers the risk for airborne infections and, in turn, the need for pesticides. In some embodiments of indoor farms, there may be a remote growing space with natural light (such as the rooftop of the warehouse). In these cases, the AS/R software, in combination with the storage rovers and forklifts, can circulate the pods and distribute the natural light optimally to the different crops.

[00179] Indoor farms are the most suitable for difficult climate conditions, where the difference between the outside climate conditions and those needed inside the farms are considerably different. In these locations, the optimal isolation will considerably decrease the overall energy consumption, as less heat or cold will leak outside.

[00180] Greenhouses offer limited climate control and air filtration but are considerably cheaper to install. The level of climate isolation in the greenhouse is adjustable based on the materials chosen. In some embodiments, the greenhouse may be covered with a net alone. This allows excellent airflow but eliminates the possibility of changing the internal climate conditions. Greenhouse farms are mainly suitable for locations where the external climate conditions are already optional for crop growth, so little change is required. In greenhouse embodiments, the growing protocols will only be able to change the nutrient composition to aid the crops when difficult climate conditions occur. If an airborne infection is discovered, pesticides must be used, as air filtration is not an option.

[00181] Greenhouse farms allow most of the required natural light to enter, thus lowering the overall energy consumption of the artificial lights. In some embodiments, the pods may not be stacked (or stacked to a low number of pods), thus allowing the sunlight to reach most of the growing crops and almost eliminating the need for artificial light.

[00182] Glasshouses are a preferred embodiment for the farm location. Glasshouses are a cross between indoor and greenhouse farms. They provide full insulation to the external air, thus allowing high levels of air filtration. As glasshouses are transparent, most of the sunlight enters the farms and reaches the growing crops. This decreases the need for artificial light. The climate is fully controllable, although a considerable amount of energy will be required if the difference in climate temperature is too wide between outside and within the farm. [00183] Most plants' physiological processes are affected by plant and ambient temperatures. For optimal control of the growing protocols, and to make the system climate-agnostic, some embodiments of the discloses system may include monitoring and controlling of the air temperature. Controlling the air temperature can be achieved in different embodiments. The optimal choice is dependent on the local energy costs and the difference between the local weather and the optimal crop temperature. An electric air conditioning system may be used in some embodiments. In warm climates, underground heat pumps may be used to lower the air temperature. This embodiment uses a deep underground tunnel and fans to push air through the tunnel. As the temperatures are constantly lower deep underground, this relatively simple and energy- efficient system can cool the air temperature within the farm. Another embodiment of underground cooling is using water (for example, the water within the nutrient tanks), which is pushed underground, cooled down, and reused. In other embodiments, radiators, which passively cool via the external airflow, may be integrated to further reduce the required energy consumption.

[00184] Monitoring and controlling airflow (the movement of air at a current set speed) may be required for optimal crop growth. Insufficient air current speed around the plants suppresses gas diffusion in the leaves, which subsequently reduces the rates of photosynthesis and transpiration and hence plant growth. Airflow also helps prevent condensation on the leaves and other surfaces, helping to prevent unwanted growth of arterial and molds. To achieve optimal airflow, mechanical fans may be used in some embodiments to generate airflow and control the current speed. Strategically positioning these fans is important for efficient use. The cloud software, based on the structure and pod parking plan, may, in some embodiments, recommend the optimal locations based on airflow simulations. In some embodiments, fans may be controlled remotely to allow dynamic timing and sequences based on the optional grow protocol of the crops.

[00185] Healthy, actively growing plants can transpire a lot of water, resulting in a rapid increase in the water vapor content and humidity inside the farm. When the air conditioning system is operating, humidity is kept under control because water vapor condenses on the cooling coils, dropping the moisture content, and thus humidity, of the air. Therefore, one embodiment for controlling humidity in the farm is to alternate the operation of lights to generate heat and cause the air conditioner to run, resulting in simultaneous cooling and dehumidification of the space. In alternative embodiments, stand-alone dehumidifiers can be installed that do not rely on the operation of air conditioners. These units may be used in applications that require distinct day/night cycles when turning on the lights for dehumidification would be undesirable. They can also be used to avoid operating lights and air conditioners during peak energy-use periods, lowering energy costs.

[00186] In some embodiments, the condensation can be significant, and recycling the condensation is desirable for conserving water. In this embodiment, the water vapor is sent for filtration before being used to prepare the nutrient solution by the nutrient management system and device.

[00187] Plants assimilate O2 during photosynthesis and release CO2 during respiration. Small changes in the CO2 concentration can have a significant impact on the rate of photosynthesis. In some embodiments, CO2 is only monitored and can be affected by other climate systems such as air conditioners or air filtration. In a preferred embodiment, CO2 is both monitored and controlled. This allows the dynamic growing protocols to set the optimal CO2 levels, based on the crops and their stage of growth.

[00188] In embodiments of the disclosed growing system, some environmental sensors are installed on each pod 100 and/or storage rover 200 It may be expensive to install all of the required sensors on each pod. Therefore, in a preferred embodiment, the more expensive sensors are installed in fixed locations throughout the farm. These sensors provide the cloud software with readings, which allows the software to estimate the overall farm climate. In some embodiments, the optional installation location for each of these sensors may be recommended by the cloud software based on the floor and structure plan.

[00189] There are several types of central sensors which may be be installed. These include non-dispersive infrared (NDIR) sensors that may be installed throughout the farm to monitor the CO2 levels; airflow (air current speed) sensors that may be used to monitor the airflow and, in turn, trigger the software to activate the mechanical fans; humidity sensors that may monitor humidity deficit; oxygen sensors that may be used to monitor and control the level of oxygen within the farm; air composition sensors that may be used to identify airborne bacteria; light spectrometers that may be used for spectrum analysis of the natural light present within the farm (the sensor data, in combination with the cloud software, may change the artificial light regime within the farm); movement sensors that may be used to identify the presence of humans within the farm. These sensors can then trigger the software to stop the robotic movements or change the CO2 levels within the farm for the safety of the humans inside the farm. Radar transmitting beacons may also be installed to assist the storage rovers in calculating their precise location within the farm.

[00190] It should be noted that, although the pods 100 are described as being vertically stacked, stacking the pods is not a requirement. Some embodiments may include pods 100 that are not vertically stacked. Instead, these pods may be arranged in a single layer (i.e., unstacked pods) in a horizontal plane. Rovers 200 may be used to transport the pods 100 between different locations for automated storage. As explained previously, the rovers 200 may travel to a location below a pod 100, lift the pod (using its jack), and transport the pod 100 to a different location (e.g., near a CNC machine for automated tending, near a farmer, etc.). The rovers 200 may be used to drive around acres of greenhouses to fetch the pods without expending human labor. Since the rovers 200 can travel under the pods 100, the pods can be arranged with a minimum spacing (or no spacing) between them. Such an embodiment may be advantageous in locations where it may be more economical to build more greenhouses instead of stacking the pods (when weighing between artificial lighting energy cost or a single level greenhouse). In some embodiments, artificial lighting may be installed on a frame that can be stacked and manipulated by rovers 200 and/or forklifts 300.

[00191] Furthermore, other areas of art may benefit from this method, and adjustments to the design are anticipated. Thus, the scope of the disclosure should be determined by the appended claims and their legal equivalents, rather than by the examples given.