TECHNICAL FIELD

[001] The present invention relates in general to sensor network systems. More particularly, the present invention relates to monitoring quality of a commodity in a plurality of process phases related with production and distribution of the commodity.

BACKGROUND

[002] With increased focus on quality and traceability in food and beverage industries, there is growing requirement for sensors for quality monitoring. Typically, the quality monitoring for a commodity from such industries is required for every sub system such as, from farm, logistics/transportation, factory to the end customers. This requires large number of sensors communicating with each other or within a network for every subsystem. For example, monitoring sensors required in farms for soil quality, animal health and nutrition, moisture, nutrients, weather conditions and the like. Quality monitoring sensors required in logistics and supply chain for tracking, cold chain status, product quality and the like and process monitoring sensors at processing factory etc.

[003] While online monitoring sensors are required for real time quality control, there are limitations with respect to availability of online quality sensors, both due to cost as well as maintenance related concerns. A common practice in such cases is use of online secondary sensor measurements and offline quality measurements to develop on-line soft sensors for quality measurements. However, availability of offline sensors, internet connectivity and sensor faults may pose problems in availability of online data.

[004] Generally, manufacturing to distribution of commodities requires a number of discrete processes to obtain an end product. The various processes, starting from initial raw material in farm, processing, supplying, transporting and packaging final end product, are very different from one another and are performed in different locations with different control systems. This ensures a flow of data from each process stage in downstream direction to subsequent processes. Thus, information available at each process stage becomes crucial for decision making for subsequent processes. [005] While most of existing systems are focusing on sensor as well as data management within process sensor network, there are no systems currently that focusses on an end-to-end user interactive sensor network which analyses feasibility of quality prediction using available measurements across sensor network and accordingly trigger offline measurement in real time. In addition, the existing systems do not provide any suggestion/actions for upstream processes. This may lead to inefficient decision making and routing of commodity to process stages based on quality. Hence, identification of actions/suggestions for upstream processes is desired to ensure desired quality and traceability of commodity.

SUMMARY

[006] The present invention relates to a method and a sensor network automation system for monitoring quality of a commodity in a plurality of process phases related with production and distribution of the commodity. A commodity may be any good or material, for example, agricultural products, fuels, metals and the like. The commodity may be processed at different industries depending on type such as, dairy industry, beverage industry, metal industry, agriculture industry and the like. The commodity undergoes a plurality of process phases for manufacturing and distribution. The plurality of process phases has a respective sensor network that monitors online and offline process and quality parameters.

[007] The method of the present invention is implemented by the sensor network automation system. The sensor network automation system can be a control system in the industries. The sensor network automation system is connected to a plurality of sensor networks.

[008] The method comprises obtaining measured data for the commodity at each of the plurality of process phase from a plurality of sensors associated with each sensor network. The commodity is identified with a virtual tag associated with processing of the commodity during the plurality of process phase and includes the measured data associated with each process phase of the plurality of process phase. In an embodiment, the commodity may be a food and beverage product. The method includes calculating values of one or more quality parameters identified for the commodity at each process phase using information related to upstream process phases from the virtual tag. The virtual tag is updated during the processing of the commodity throughout the process phase. In an embodiment, the information updated in the virtual tag can be retrieved using an ID mark associated with the commodity. Further, the method includes determining quality of the commodity at end of each process phase. The updated virtual tag is compared with threshold of quality parameters at each process phase using a predefined model.

[009] As a result of the comparison, the method comprises enabling at least one action in at least one upstream process phase.

[010] In accordance with different embodiments, the sensor network automation system comprises a processor, a network interface and a memory communicatively coupled to the processor. The network interface may obtain measured data for the commodity at each of the plurality of process phase from the plurality of sensors associated with each sensor network. The commodity is identified with a virtual tag associated with processing of the commodity during the plurality of process phase and includes the measured data associated with each process phase of the plurality of process phase. The processor calculates values of one or more quality parameters identified for the commodity at each process phase using information related to upstream process phases from the virtual tag. The virtual tag is updated during the processing of the commodity throughout the process phase. Further, the processor determines quality of the commodity at end of each process phase and compares the updated virtual tag with threshold of quality parameters at each process phase using a predefined model. Based on result of the comparison, the processor enables at least one action in at least one upstream process phase.

BRIEF DESCRIPTION OF DRAWINGS

[Oi l] Figure 1 shows an environment for monitoring quality of a commodity in a plurality of process phases during production and distribution, in accordance with an embodiment of the invention;

[012] Figure 2 is a simplified block diagram of a sensor network automation system for monitoring quality of a commodity in a plurality of process phases, in accordance with an embodiment of the invention;

[013] Figure 3 shows an exemplary embodiment for monitoring quality of a commodity in a plurality of process phases during production and distribution, in accordance with an embodiment of the invention; and [014] Figure 4 is a flowchart of a method for monitoring quality of a commodity in a plurality of process phases during production and distribution, in accordance with an embodiment of the invention.

DETAILED DESCRIPTION

[015] A commodity undergoes multiple processing phases before reaching to final customers. Industries such as, process industries, process the commodity and provide details associated with the processing to subsequent process phases.

[016] Generally, manufacturing up to distribution of the commodity requires a number of processes to obtain an end product. The various processes, starting from initial raw material in farm, processing, supplying, transporting and packaging final end product are very different from one another and are performed in different locations. Typically, the entire process phase ensures a flow of commodity from each process stage in downstream direction to subsequent process phase. Thus, information available at each process stage becomes crucial for decision making for subsequent processes as any error or changes in data arriving from upstream process phase may impact the commodity in the subsequent process phase. Currently, available sensor networks focus on monitoring and providing inputs to process phases in downstream direction. However, there is scope of improving downstream processes by influencing operations at upstream phases in real time. The present invention provides a method and a sensor network automation system for monitoring quality of the commodity in such environment.

[017] Figure 1 shows an exemplary environment 100 of a process industry which comprises a sensor network automation system 101 for monitoring quality of a commodity in a plurality of process phases related with production and distribution of the commodity. The commodity may refer to any goods, material or product for example, agricultural products, fuels, dairy product, metals and the like. In an embodiment, the commodity may include a food or beverage. The sensor network automation system 101 is connected to a sensor network 103i, a sensor network 1032, . and a sensor network 103N (collectively referred as plurality of sensor networks 103). Each of the plurality of sensor network 103 may be located at different locations and is associated with a process phase of a plurality of process phases.

[018] The plurality of process phases may be associated with processing, manufacturing, logistics, factory, transportation, customer review and the like. The plurality of process phases may differ depending on the type of commodity. For example, when the commodity is a dairy product, the plurality of process phases may include, cattle farming, collection of milk, transportation, processing, logistics, storage and the like. Likewise, in an agriculture process, the plurality of process phases may include, crop selection, land preparation, seed selection and sowing, irrigation, crow growth, fertilization, transportation and the like.

[019] In order to process the commodity, each of the sensor network includes a plurality of sensors. As shown in Figure 1, the sensor network 103i includes a sensor 105n, a sensor 105i2,..and a sensor 105IN. The sensor network 1052 includes a sensor 10521, a sensor

10522. and a sensor 1052N. Likewise, each of the sensor network includes respective sensors (which are collectively referred as plurality of sensors 105).

[020] At beginning of the processing, a virtual tag is created for each commodity. The virtual tag may refer to a unique tag which includes entire history related to the processing of the commodity. When the processing of the commodity is initiated, the sensor network automation system 101 may obtain measured data for the commodity at each of the plurality of process phase from the plurality of sensors 105 associated with each sensor network 103. In case of unavailability of the measured data, the sensor network automation system 101 may trigger offline measurements for the plurality of sensors 105 using a Human Machine Interface (HMI) located at each of the plurality of sensor networks 103.

[021] Each of the plurality of process phase may include the processing information in the virtual tag associated with the commodity. Thus, the virtual tag of the commodity includes the measured data associated with each process phase of the plurality of process phase. For example, in the agriculture industry, measured data regarding soil such as, soil bacterial load, pesticide content and the like may be stored in the virtual tag. On receiving the measured data, the sensor network automation system 101 may calculate values of one or more quality parameters identified for the commodity at each process phase. In an embodiment, the one or more quality parameters for the commodity may be identified using any existing known techniques depending upon the industry.

[022] The one or more quality parameters may differ from one commodity to another. For example, the one or more quality parameters for beverage industry, such as for beer production may be appearance, aroma, PH level, flavour, color and the like. Particularly, the sensor network automation system 101 may calculate the one or more quality parameters using information provided by one or more upstream process phases in the virtual tag. In an embodiment, the one or more upstream process phases may refer to foregoing or preceding process phases with respect to a process phase. While processing the commodity, the virtual tag is updated by each of the plurality of sensor network 103 throughout the process phase. The information updated in the virtual tag is retrieved at each of the plurality of process phase using a unique Identification (ID) mark associated with the commodity. The ID mark associated with the commodity may be provided to operators at respective location of the sensor network.

[023] Further, the sensor network automation system 101 may determine a quality of the commodity at end of each process phase. In an embodiment, the quality of the commodity may be determined using any known existing techniques.

[024] Upon determining the quality of the commodity, the sensor network automation system 101 may compare the updated virtual tag with threshold of quality parameters at each process phase using a predefined model (not shown in Figure 1, covered in Figure.2) configured at the sensor network automation system 101. The predefined model may be any statistical model that is generated offline based on standard quality data associated with different commodities. In an embodiment, the model can be generated using various techniques such as, using one or more of machine learning, filtering and so forth. In an embodiment, the model is trained using deep learning techniques such as Convolutional Neural Networks (CNN) with standard quality data of the commodity.

[025] Thereafter, based on a result of the comparison, the sensor network automation system 101 may enable at least one action in at least one upstream process phase. The at least one action may be utilised in the at least one upstream process phase for a next batch of process for the same commodity. The one or more action may include rescheduling the production of the commodity based on predicted disturbances determined based on the comparison.

[026] An exemplary embodiment for monitoring quality of a dairy commodity during production and distribution is illustrated in Figure.3.

[027] Figure.3 shows four process phase for a commodity in diary industry such as, milk. The process phase includes cattle farming as phase 1, collection and milk transportation as phase 2, processing as phase 3 and logistic and storage as phase 4. The phase 1 involves various aspects related to health and productivity of cattle. The phase 1 is performed at farm location and includes sensors such as, respiration sensor, humidity sensor, rumination sensor and the like, which are used in advanced cattle health monitoring system. Further, the cattle farm may also include sensors related to disease detection and tracking and may include, temperature, heartbeat and piezoelectric sensors for detecting weakness in the cattle. In an embodiment, the one or more sensors in the cattle farm are in place and are used to monitor health of the cattle, which directly affects milk yield and quality. In an embodiment, some of these sensors may be configured for online monitoring. Whereas certain parameters may require offline measurements such as, lab tests (blood test, urine test, etc.) to measure physiological parameters affecting health of the cattle. The information associated with the processing in the phase 1 is updated in the virtual tag for the milk commodity.

[028] The phase 2 includes measurements such as, cow milk conductivity, yield, temperature, bacterial load, milk let-down flow rate in near real time and the like. The phase 2 may include various temperature sensors for monitoring temperatures during transportation. In an embodiment, the one of more such measurements in phase 2 can be online or offline based on lab analysis of the commodity. The transportation sensor network for phase 2 may utilise the information from the virtual tag updated at the phase 1.

[029] The phase 3 includes various sensors such as, temperature sensors, pressure sensors, flow sensors and pH and nitrogen sensors to determine the processing as well as the milk qualities such as, protein and fat content and bacterial load. Further, the phase 4 may include cold chain temperature monitoring and microbial content monitoring. Consider in the phase 3, based on the virtual tag, the sensor network automation system 101 may identify temperature issues during transportation at specific time, which may result in spoiling the milk. In such case, the sensor network automation system 101 may update the virtual tag and may enable actions to maintain the temperature at specific degrees for next batch of the milk production.

[030] Figure 2 is a simplified block diagram of a sensor network automation system for monitoring quality of a commodity in a plurality of process phases, in accordance with an embodiment of the invention. As shown, the sensor network automation system 101 comprises a memory 201, a processor 203 and a network interface 205. The memory may include a plurality of virtual tags 207 associated with different commodities and a model 209.

[031] The network interface may include I/O’s of the sensor network automation system 101, which connect the sensor network automation system 101 with the plurality of sensor networks 103. Thus, the sensor network automation system 101 communicates with other components through the I/O’s. The sensor network automation system 101 receives the measured data from the plurality of sensors 105 associated with each sensor network through the I/O’s. The measured data associated with the commodity is captured while manufacturing and processing the commodity at the plurality of process phase.

[032] The memory 201 includes the plurality of virtual tags 207 associated with plurality of commodities and the pretrained model 209 (refer description of Figure 1 above).

[033] The processor 203 is configured to obtain the measured data for the commodity from the network interface 205. The processor 203 receives the virtual tag from each of plurality of process phase. The virtual tag includes the measured data associated with each process phase of the plurality of process phase.

[034] On obtaining the measured data, the processor 203 may calculate values of the one or more quality parameters for the commodity at each process phase. The processor 203 calculates the values by using the information provided from the upstream process phases in the virtual tag. For example, the quality parameter such as, bacterial load of milk, at processing phase, may be calculated based on the temperature information from the preceding transportation phase.

[035] Thus, the processor 203 continuously receives the measured data at different process phase, which is used for updating the virtual tag. Further, the processor 203 determines the quality of the commodity at end of each process phase. The quality of the commodity may be determined using any existing known techniques. Based on the quality, the processor 203 may compare the updated virtual tag with threshold of quality parameters at each process phase. In an embodiment, the threshold of quality parameters may be predetermined based on standard quality parameters. The processor may compare the updated virtual tag using the model 209.

[036] Thereafter, the processor 203 may enable at least one action in the at least one upstream process phase based on the result of the comparison.

[037] The above Figures 1 - 3 are explained considering the dairy industry. The present invention is not restricted to dairy commodity or dairy industry. The present invention can also be implemented in other processing plants such agriculture industry, beverage industry and the like.

[038] Additionally, the present invention can also be implemented in other industries providing services.

[039] Referring now to Figure 4, which is a flowchart of a method for monitoring quality of a commodity in a plurality of process phases during production and distribution, in accordance with an embodiment of the invention.

[040] Various steps of the method may be performed by the sensor network automation system 101 , or at least in part by the sensor network automation system 101.

[041] At 401, the measured data for the commodity at each of the plurality of process phase is obtained by the network interface 205 from the plurality of sensors 105 associated with each sensor network. The commodity is identified with the virtual tag associated with processing of the commodity during the plurality of process phase and includes the measured data associated with each process phase of the plurality of process phase.

[042] At block 402, the values of the one or more quality parameters for the commodity at each process phase is calculated by the processor 203 using the information related to upstream process phases from the virtual tag.

[043] At 403, the virtual tag is updated by the processor 203 during the processing of the commodity throughout the process phase.

[044] At 404, the quality of the commodity is determined by the processor 203 at end of each process phase. The quality of the commodity may be determined depending on the type of process phase. In an embodiment, the quality may be determined using any existing techniques.

[045] At 405, the updated virtual tag is compared with threshold of quality parameters by the processor 203 at each process phase using the predefined model. The predefined model includes previously available measurements (online and offline) for the commodity.

[046] At block 406, at least one action is enabled in at least one upstream process phase based on the result of the comparison. The action can be implemented for the at least one upstream process during manufacture and production of next batch of process associated with the commodity.

[047] The present invention provides an improved and proactive quality control mechanism using available network of on-line/off-line sensors.

[048] The present invention enables routing flow of commodity to an appropriate/ideal downstream entity to maximize gain for the quality of the commodity at any specific point of time. For example, the cattle farmer can route raw material to an appropriate collection centre, or a collection centre can route material to an appropriate processing centre and so on to choose a scenario that favors better utilization of the quality of milk at any point of time.

[049] The present invention enables to plan a better resource planning in line with requirements of flexible and evolving sensor network. For example, based on data related to external environment from past, productivity and efficiency of any processing phase, customer feedback and the like, the resource planning of the entire end to end process phase can be optimized. That is, for instance, a farmer can plan on livestock required to maintain, a collection centre can plan on the logistics that is intended to put in place, and so on, in line with anticipated external influence.

[050] The present invention provides flexibility to negotiate pricing based on real time market dynamics at every process stage. This may ensure that at every process stage, the pricing is based on the quality and availability of the commodity and in line with requirements.

[051] The present invention in an embodiment may eliminate the need for having a fixed expiry date for the commodity in transit (including to the customer). This may facilitate better utilization across the process phase.

[052] The present invention may avoid resource saving in various steps of the process phase and accelerates faster movement of the commodity.

[053] The present invention triggers periodic monitoring of quality measurements offline and validates the model accuracy.

[054] The present invention facilitates consistent quality monitoring and control for all stakeholders. [055] In case of non-availability of online measurements, the present invention may trigger offline measurements using a Human Machine Interface (HMI) at various locations in sensor network and accordingly manage the process phase downstream to minimize production loss.

[056] The present invention triggers analytics and decision support to facilitate maximum value addition in the entire process phase to reschedule production based on predicted and/or measured disturbances (change in requirements, quality deviations, resource constraints, etc.) [057] The present invention provides convenience for the users in the process phase to monitor on-demand off-line measurements based on the trigger from the system.

REFERRAL NUMERALS