FIELD OF THE INVENTION

The invention relates to a lighting device for illuminating an environment. The invention further relates to a method and a computer program product for controlling a lighting device for illuminating an environment.

BACKGROUND

Smart systems are a burgeoning market. These systems are built around smart devices and home networks. Smart systems are often connected to the Internet, typically such that smart devices placed within an environment may be controlled, e.g. by a user when (s)he is out-of-home. The connected smart devices may be any devices capable of being connected to, or identified by, the system. Commonly used phrases for such a system and its devices is the Internet of Things (IoT) and IoT devices.

An example of a smart system is a connected lighting system, which refers to a system comprising one or more lighting devices which are controlled not by (or not only by) a traditional wired, electrical on-off or dimmer circuit, but rather by using a data

communications protocol via a wired or more often wireless connection, e.g. a wired or wireless network. Typically, the lighting devices, or even individual lamps within a lighting device, may each be equipped with a wireless receiver or transceiver for receiving lighting control commands from a lighting control device according to a wireless networking protocol such as Zigbee, Wi-Fi or Bluetooth.

In such a connected lighting system, there are still many unconnected standalone lighting devices. One of the reasons for such standalone devices/systems is the cost involved with a total connected system, e.g. costs with communication transceivers, gateways, sensors, user interface devices or wall panels. The standalone lighting devices are controlled manually by a user or based on predefined light settings, for example, based on a time of day or based on an activity of the user.

US 2016/295658A1 discloses a lighting unit (10) that utilizes sensor data to select a fade-in or fade-out profile for one or more light sources (12). The lighting unit includes a clock module (20), an ambient light level sensor (18), and a controller (16). The controller receives ambient light level data from the ambient light level sensor and time of day data from the clock module, and utilizes the received information to automatically select a fade-in or fade-out profile. The controller can also use the received ambient light level data to calibrate the clock module.

SUMMARY OF THE INVENTION

The inventors have realized that controlling a standalone lighting device based on a current time of day requires an estimation of the current time. The inventors have further realized that the accurate estimation of the current time of day either requires an accurate internal clock or an external network connection. Having an accurate internal clock is an expensive option, while network connectivity may not be viable for a standalone lighting device.

It is therefore an object of the present invention to provide a lighting device for illuminating an environment which can accurately estimate a current time of day without the need of an expensive accurate internal clock and/or a connection with an external network. The lighting device may be then controlled based on the estimated current time of day.

According to a first aspect, the object is achieved by a lighting device for illuminating an environment, comprising: a light source; a memory arranged for storing a light profile; wherein the light profile comprises lighting control instructions for controlling a characteristic of the light source based on a time of day; a processor comprising: an input unit arranged for obtaining operational state data of the lighting device; wherein the operational state data comprises information indicative of durations of one or more previous and current operational states; a processing unit arranged for inferring, by a trained machine, the current time of day based on the obtained operational state data; wherein the trained machine has been trained using a training data set comprising a switching profile and wherein the switching profile comprises the operational states with respect to the time of day; wherein the processing unit is further arranged for accessing the light profile from the memory; and a controller arranged for controlling the lighting device based on the inferred current time of day and based on the light profile.

The lighting device is arranged for illuminating an environment, for example, an indoor environment such as an office, educational institution, shopping/groceries store, residential complex, hospital, factory, etc. The lighting device comprises a memory to store a light profile, wherein the light profile comprises lighting control instructions for controlling a characteristic of the light source based on a time of day.

The lighting device further comprises an input unit arranged for obtaining the operational state data of the lighting device. The operational state data comprises information indicative of durations of one or more previous and current operational states. The operational states may be the operational states of the light source of the lighting device, for example, an operational state may comprise an On state or an Off state of the light source. In the On state, the lighting device illuminates the environment and in Off state the lighting device does not illuminate the environment. In both states, power may be provided to the lighting device. The processing unit is further arranged for inferring, by a trained machine, the current time of day based on the obtained operational state data; wherein the trained machine has been trained using a training data set comprising a switching profile and wherein the switching profile comprises the operational states with respect to the time of day. The training of the model may comprise constructing algorithms which can leam from the training data set and preferably make predictions based on input data, for instance in this case the input data may be the operational state data. The training data set may be a set of examples used to fit parameters, for instance weights, of the trained machine.

Since the processing unit of the lighting device can infer the current time of day based on the operational state data, and a controller of the lighting device is arranged for controlling the lighting device based on the inferred current time of day and based on the accessed light profile, a lighting device is provided which can estimate a current time of day without the need of an expensive accurate internal clock and/or a connection with an external network. The lighting device may be then controlled based on the estimated current time of day.

The lighting device may be connectable to an external network, and the processing unit may be further arranged for: determining whether the lighting device has been connected to the external network; and receiving the current time of day from the external network.

When the lighting device may be connectable to an external network, for example when the lighting device is comprised in a connected lighting system, the processing unit may be arranged for determining whether the lighting device has been connected to the external network, e.g. to internet, and receiving the current time of day from the external network. The light profile may be a circadian light profile. In this embodiment, the light profile may correspond to a circadian rhythm. A circadian rhythm is any biological process that displays an endogenous, entrainable oscillation of about 24 hours. These 24-hour rhythms are driven by a circadian clock, and they have been widely observed in plants, animals, fungi, and cyanobacteria. Light effects on circadian rhythm are the effects that light has on circadian rhythm.

In an embodiment, a change in the operational state may be caused by a user input. In an example, the lighting device is a standalone device and the operational state of the lighting source of the lighting device may be changed by the user input. The user input may be, for instance, in the form a device such as legacy wall switch or a switch on the lighting device. The user input may be in the form of a user activity.

The lighting device may be comprised in a lighting system, and wherein the input unit may be arranged for obtaining the switching profile from the lighting system and the processing unit may be further arranged for providing the trained machine by training an untrained machine based on the obtained switching profile.

The lighting device may be comprised in a lighting system, for example a connected lighting system, or alternatively it may a non- or a partially connected lighting system. In this case, the input unit may obtain the switching profile from the lighting system. For instance, the switching profile may be obtained from another lighting device in the lighting system.

The processing unit may be further arranged for: obtaining information indicative of environmental characteristics of the environment; and selecting the switching profile based on the environmental characteristics.

The lighting system may be comprised in the environment, and the processing unit may be further arranged for: obtaining information indicative of environmental characteristics of the environment. For example, the characteristic may be a type of environment, for instance an office, a factory, a home, a grocery store or a hospital.

Furthermore, based on the characteristic, e.g. type, the processing unit may be arranged for selecting the switching profile. For example, when the lighting device is placed in an office, the processing unit selects a switching profile of an office (may be same office or from another office).

The memory of the lighting device may be further arranged for storing the switching profile of the lighting device and the processing unit may be further arranged for providing the trained machine by training an untrained machine based on the stored switching profile.

Alternative to obtaining the switching profile from the lighting system, the switching profile of the lighting device may be stored in the memory and the processing unit may be further arranged for providing the trained machine by training an untrained machine based on the stored switching profile.

The training data set may comprise previous operational state data; and the processing unit may be further arranged for providing the trained machine by training the untrained machine based on the previous operational state data. In another embodiment, the switching profile may comprise the previous operational state data.

The processing unit may use previous operational state data for training the untrained machine; wherein the previous operational state data may comprise information indicative of durations of one or more previous operational states. In another embodiment, the switching profile may comprise the operational state data and the processing unit may be arranged for training the untrained machine using switching profile which may also comprise previous operational state data.

If the the amount of information in the training data set is below a threshold, the controller may be arranged for controlling the lighting device according to predefined light settings.

In case when the information content in the training data set is below a certain threshold, the training of the untrained machine may not be feasible. In such cases, the predefined light settings may be used by the controller to control the lighting device.

The training of the untrained machine may comprise using machine learning algorithms.

One of the methods to train the untrained machine may be a machine learning algorithm, such as a support vector machine, a deep learning algorithm, etc. In a simple embodiment, training the untrained machine may comprise comparing the switching profile with the operational state data and inferring the time based on said comparison. In a more advance embodiment, deep learning methods, e.g. using artificial neural networks, may be used as training algorithms.

The training data set may further comprise a second set of features, wherein the second set of features may comprise one or more of: a weather condition, a measure of ambient light, an occupancy measure and an activity of the user; and the processing unit may be further arranged for inferring, by the trained machine, the current time of day based on the second set of features.

A second set of features comprising weather condition, an activity of user, etc. may be used as a training data set. For example, the activity of a user in a factory may be a good candidate for the training data set. In this case, the processing unit may be further arranged for inferring, by the trained machine, the current time of day based on the second set of features. The processing unit may use the second set of features or the switching profile or both the second set of features and the switching profile as training data set.

In an embodiment, the characteristic of the lighting device may comprise one or more of: color, color temperature, intensity, beam width, beam direction, illumination intensity, and/or other parameters of one or more of the light sources of the lighting device.

The controller may be arranged for controlling a characteristic of the lighting device, wherein the characteristic may comprise color, intensity, etc. An example of changing, e.g. the color is by using the Tunable white technology which is based on the natural color changes in the light over the course of the day.

According to a second aspect, the object is achieved by a method of controlling a lighting device for illuminating an environment, the method comprising:

obtaining operational state data of the lighting device; wherein the operational state data comprises information indicative of durations of one or more previous and current operational states; inferring, by a trained machine, a current time of day based on the obtained operational state data; wherein the trained machine has been trained using a training data set comprising a switching profile and wherein the switching profile comprises the operational states with respect to the time of day; accessing a light profile from a memory of the lighting device, wherein the light profile comprises lighting control instructions for controlling a characteristic of a light source of the lighting device based on the time of day; controlling the lighting device based on the light profile.

According to a third aspect, the object is achieved by a computer program product comprising instructions configured to cause a lighting device according to the first aspect to execute the steps of the method according to the second aspect.

It should be understood that the computer program product and the system may have similar and/or identical embodiments and advantages as the above-mentioned methods. BRIEF DESCRIPTION OF THE DRAWINGS

The above, as well as additional objects, features and advantages of the disclosed devices and methods will be better understood through the following illustrative and non-limiting detailed description of embodiments of devices and methods, with reference to the appended drawings, in which:

Fig. 1 shows schematically and exemplary an embodiment of a system with lighting device(s) for illuminating an environment;

Fig. 2 shows schematically and exemplary an embodiment of a lighting device for illuminating an environment;

Fig. 3 shows schematically and exemplary a flowchart illustrating a method of controlling a lighting device for illuminating an environment;

Fig. 4 shows schematically and exemplary an embodiment of a lighting device for illuminating an environment.

All the figures are schematic, not necessarily to scale, and generally only show parts which are necessary in order to elucidate the invention, wherein other parts may be omitted or merely suggested.

DETAILED DESCRIPTION OF EMBODIMENTS

A circadian rhythm is any biological process that displays an endogenous, entrainable oscillation of about 24 hours. Entrainment occurs when rhythmic physiological or behavioral events match their period to that of an environmental oscillation. Entrainment is ultimately the interaction between circadian rhythms and the environment. These 24-hour rhythms are driven by a circadian clock, and they have been widely observed in plants, animals, fungi, and cyanobacteria. At all times of the day, we are supposed to have a natural rhythm that guides our alertness, sleeping, and eating patterns.

Although circadian rhythms are endogenous, i.e. they are "built-in" or self- sustained, they can be adjusted (entrained) to the local environment by external cues called zeitgebers. Light, besides other zeitgebers such as temperature and redox cycles, has a strong impact on circadian rhythm. Studies of the circadian rhythm have shown that the intensity, timing, duration, and wavelength of light can affect the biological clock of humans. For example, researchers have found that CCTs (Correlated Color Temperature) above 5500K (containing a high amount of blue light) can disrupt our circadian rhythm by keeping us up when we should naturally be sleeping. In the daytime when we need to focus, the lighting with CCT of 6000K helps in concentration and in the evening before bed, the CCT can be decreased to 3000K, promoting the natural release of melatonin to help you sleep.

Controlling a lighting device based on a current time of day requires an accurate knowledge of the current time of day. In a connected lighting system, where the lighting devices are (permanently) connected to an external network, e.g. the internet, the current time of day may be obtained from the external network. Alternative to the network connectivity, the lighting device may be equipped with an accurate internal clock for obtaining the current time of day. All these solutions are expensive and/or cumbersome. Specifically, for a standalone lighting device, the connectivity may not be a viable option and having an internal clock may be expensive. All known implementations, e.g. indoor, include some form of clock synchronization, which are usually expensive.

Fig. 1 shows schematically and exemplary an embodiment of a system 100 with lighting device(s) 1 lOa-d for illuminating an environment 101. The system exemplary comprises four lighting devices 1 lOa-d. A lighting device is a device or structure arranged to emit light suitable for illuminating an environment, providing or substantially contributing to the illumination on a scale adequate for that purpose. A lighting device comprises at least one light source or lamp, such as an LED-based lamp, gas-discharge lamp or filament bulb, etc., optionally with any associated support, casing or other such housing. Each of the lighting devices may take any of a variety of forms, e.g. a ceiling mounted lighting device, a wall- mounted lighting device, a wall washer, or a free-standing lighting device (and the lighting devices need not necessarily all be of the same type). In this exemplary figure, the lighting devices 1 lOa-c are ceiling mounted and the lighting device 1 lOd is a free-standing lighting device. The system 100 may contain any number/type of the lighting devices l lOa-d.

The lighting devices 1 lOa-d are arranged for illuminating the environment 101. The environment may be an office, a factory, a house, a grocery store or a hospital. The system may further comprise a sensor 140. The sensor may for example be a motion sensor to detect motion/presence, a light sensor for detecting ambient light levels, a temperature sensor, a humidity sensor, a gas sensor such as a C02 sensor, a particle measurement sensor, an audio sensor and an imaging sensor such as a camera. The presence detection sensor may be a passive infrared sensor or active ultrasound sensor. The system may further comprise a wall-switch 130 which may be arranged for changing the operational state of the lighting devices l lOa-d. The sensor 140 may be used to change the operational state of the lighting device l lOa-d. The lighting devices 1 lOa-d may be comprised in a lighting system. The lighting system may be a connected system, such as a connected lighting system, wherein the lighting devices 1 lOa-d are connected to an external network, e.g. internet. The lighting system may be such that only the ceiling mounted lighting devices 1 lOa-c are connected and the free-standing lighting device 1 lOd is not connectable. The lighting system may be locally connected such that the lighting devices 1 lOa-d are controlled based on the sensor 140, for instance, the lighting devices 1 lOa-d may change the operational state when the presence detection sensor 140 detects presence of a user 120. In another example, the lighting devices may be locally connected to each other, e.g. by using Zigbee or Bluetooth. The lighting system may be non-connected, meaning that all the lighting devices 1 lOa-d are standalone devices and a change in the operational state is caused by a user input, e.g. by the legacy switch 140 or a switch (not shown) comprised in each lighting device l lOa-d. There may be four different wall switches 140, one for each lighting device 1 lOa-d. Other combinations of connecting and controlling the lighitng devices 1 lOa-d are also possible.

Fig. 2 shows schematically and exemplary an embodiment of a lighting device 210. The lighting device 210 comprises a light source or lamp 211 such as an LED-based lamp, gas-discharge lamp or filament bulb, etc. The lighting device 210 further comprises a controller 215 arranged for controlling the light source 211. The controller 215 may be implemented in a unit separate from the lighting device 210, such as wall panel, desktop computer terminal, or even a portable terminal such as a laptop, tablet or smartphone. In such case, the lighting device 210 is connected to the controller 215 via wireless or wired communication. Alternatively the controller 215 may be incorporated into the same unit as the lighting device 210. Further, the controller 215 may be implemented in a single unit or in the form of distributed functionality distributed amongst multiple separate units (e.g. a distributed server comprising multiple server units at one or more geographical sites, or a distributed control function distributed amongst the lighting device 210). Furthermore, the controller 215 may be implemented in the form of software stored on a memory (comprising one or more memory devices) and arranged for execution on a processor (comprising one or more processing units), or the controller 215 may be implemented in the form of dedicated hardware circuitry, or configurable or reconfigurable circuitry such as a PGA or FPGA, or any combination of these.

The lighting device 210 further comprises a memory 212 which may be arranged for storing a light profile, a switching profile and/or an operational state data. The light profile comprises lighting control instructions for controlling a characteristic of the light source based on a time of day. The characteristic of the light source of the lighting device may comprise one or more of: color, color temperature, intensity, beam width, beam direction, illumination intensity, and/or other parameters of one or more of the light sources of the lighting device.

The light profile may be a circadian light profile. For example, in an office setting, brighter lights may be used to work better during the early hours and increase work- flow, whereas softer lights may be used to work better with concentration in the evenings. This natural transition between brighter lights in the early hours and warmer lights in the evenings reflect the natural states and thus may manipulate pyscological state of people present in the office. Similarly, in a classroom, lighting with a CCT between the 4000K and 6000K may help with alertness and improves students efficiency. In a hospital environment, the light profile may be a low-intensity warm colour temperature in the early morning, cool colour temperature in mid-moming and high-intensity in the afternoon, falling back to low- intensity warm colour temperature in the evenings, which may be used to counter the disruption to normal sleep-wake cycles experienced in hospitals. In an example, an implementation of the circadian light profile in the lighting devices 1 lOa-d may be achieved by using tunable white technology.

The operational states may be the operational state of the light source 211 of the lighting device 210. For example, operational state may comprises an on state or an off state of the light source 211. The operational state may further comprise the information indicative of durations of the operational state (On/Ofl), wherein a change in the operational state may be caused by a user input. The change may also be caused by a sensor 140 trigger, when the lighting device 210 is connected to the sensor 140, e.g when a presence sensor 140 detects a presence, it triggers the lighting device to change the operational state, e.g. from an Off state to an On state. The operational state data may comprise information indicative of durations of one or more previous and current operational states. The switching profile may comprise the operational states with respect to the time of day which is shown exemplary in figure 4.

The lighting device 210 may comprise a processor 213, wherein the processor 213 may comprise an input unit 213a and a processing unit 213b. The input unit 213a may be arranged for obtaining the operational state data of the lighting device 210. For example, the input unit 213a may obtain the operational state data from the memory 212 or may obtain it from an external device via a wireless or a wired connection via an interface 214. Any other source of obtaining the operational state data is also possible. The processor 213 may be implemented in a unit separate from the lighting device 210, and in such case, the lighting device 210 is connected to the processor 213 via wireless or wired communication.

Alternatively the processor 213 may be incorporated into the same unit as the lighting device 210. Further, the processor 213 may be implemented in a single unit or in the form of distributed functionality distributed amongst multiple separate units. Furthermore, the processor 213 may be implemented in the form of software stored on a memory, or the processor 213 may be implemented in the form of dedicated hardware circuitry, e.g. a driver circuitry or configurable or reconfigurable circuitry such as a PGA or FPGA, or any combination of these.

The processing unit 213b may be arranged for inferring, by a trained machine, the current time of day based on the obtained operational state data; wherein the trained machine has been trained using a training data set comprising a switching profile and wherein the switching profile comprises the operational states with respect to the time of day. The lighting device 210 may further comprise an inaccurate internal clock (not shown) which may be synchronized by the inferred time of day. The inference of the time of day may be based on the inferred time and by the inaccurate internal clock.

An example of switching profile is shown in figure 4, which exemplary comprises four different switching patterns 450a-d of the lighting device 210 representing, for instance four different days of the week. For example, the lighting device is comprised in an environment, e.g. an office. A similar pattern of switching between On state and Off state of the lighting device 210 may be observed. For example, people entering the office around tl, for instance at 8am; and as a result, the operational state of the lighting device 210 is changed to an On state. The operational state switches from the On state to an Off state when people go out to for a short coffee break at t2 and return at t3. Similarly, the switching occurs at the lunch time between t4 and t5; and then may be again at another coffee break between t6 and t7. At t5, for instance at 5pm, the people leave the office. This switching profile is a very typical switching profile of an office environment. For a different environment, for example, a factory the switching profile will be different.

Now coming back to figure 2, the lighting device 210 may also comprise a communication unit (not shown), which is arranged for sending and receiving

communication signals to and from an external network (not shown), e.g. internet, or locally to and from the other lighting devices 1 lOa-d or the sensor 140. The communication signal may comprise control instructions to control the characteristic of the light source 211. If the lighting device 210 is connectable to the external network; the processing unit 213b may be further arranged for determining whether the lighting device has been connected to the external network; and receiving the current time of day from the external network. The processing unit 213b may perform the determination based on the status of the

communication unit. In an example, the lighting device 210 is connectable to the external network but the processing unit 213b may determine that the lighting device 210 is not connected to the external network, e.g. may be because of connectivity issues, then the processing unit 213b may be arranged for inferring, by a trained machine, the current time of day based on the obtained operational state data.

Fig. 3 shows schematically and exemplary a flowchart illustrating a method of controlling a lighting device for illuminating an environment. In the obtaining step 310, an operational state data of the lighting device is obtained; wherein the operational state data may comprise information indicative of durations of one or more previous and current operational states. The operational state data may be obtained from the memory 212 or from an external device via the interface 214 or may be obtained internally from other components of the lighting device 210.

A current time of day is inferred 320, by a trained machine, based on the obtained operational state data. The trained machine has been trained using a training data set comprising a switching profile and wherein the switching profile may comprise the operational states with respect to the time of day. A training dataset is a dataset of examples used for learning, that is to fit the parameters (e.g. weights) of a machine, for example, a classifier. A machine or a model is a structure and corresponding interpretation that summarizes or partially summarizes a set of data, for description or prediction. The machine or model may also summarize the relationship or mapping between an input and an output. Most inductive algorithms generate machines or models that can then be used, e.g. as classifiers, as regressors, as patterns for human consumption, and/or as input to subsequent stages of the KDD process.

When the lighting device 210 is comprised in a lighting system, the input unit 213a may be arranged for obtaining the switching profile from the lighting system. The switching profile may be obtained via a wired or wireless connection using any

communication technology known in the art. The processing unit 213b may be further arranged for providing the trained machine by training an untrained machine based on the obtained switching profile. In a simple example, the trained machine infers the current time of day by correlating or comparing the switching profile with the obtained operational state data and infer the current time of day based thereon. In another simple example, a low pass filter operation may be applied on the switching profile of the lighting device and/or obtained from the lighting system and the mid-point of the day as the peak of the outcome of low-pass filter may be determined. The actual time may then be estimated by counting mains cycles (every mains cycle is 1/50 s) with reference to the estimated mid-point. In an alternative embodiment, as an initialization step the processor 213 may receive a time of day during configuration or commissioning of the lighting system and the training of the untrained machine is continued until an accuracy above a threshold is achieved. When the lighting device 210 is comprised in the lighting system, the inferred time of day by the lighting devce 210 may be shared with other lighting devices 1 lOa-d also comprised in the lighting system. In another example, the training of the untrained machine comprises using machine learning algorithms, for instance, the trained machine is a classifier, with limited bounded set of outputs, e.g. the trained machine may only infer time after 15 or 30 min. In an advanced example, the training of the untrained machine may comprise deep learning algorithms, such as convolutional neural netowrks. The training of the untrained machine may be performed by the processing unit 213b of the lighting device 210. Alternatively, the training of the untrained machine may be performed external to the lighting device 210 and the input unit 213a may be arranged for obtainin the trained machine and further arranged for inferring the current time of day using the trained machine. Another alternative may be that the input unit 213a obtains a trained machine and the processing unit 213b retrains it using the training data set, for example, by using algorithms such as transfer learning. Any other combination and training method known in the art is also possible to train the untrained machine.

The selection of the switching profile, e.g. as shown in figure 4, for training of the untrained machine may be based on environmental characteristics. The processing unit 213b may be further arranged for: obtaining information indicative of environmental characteristics of the environment; and selecting the switching profile based on the environmental characteristics. The characteristic may be the type of environment, e.g. an office environment and or a factory environment. As the routine in an office for people coming and leaving, and therefore the switching profile, may be different compared to a factory, the selection of the switching profile is based on the einvironment. For a lighting device 210 comprised in an office, the switching profile of an office is used.

The training of the untrained machine may be based on the stored switching profile of the lighting device 210, wherein the switching profile is stored in the memory 212. The processing unit 213b may be further arranged for providing the trained machine by training an untrained machine based on the stored switching profile. As discussed before, different training methods and possibilities of using external device are also possible for training the untrained machine based on the stored switching profile of the lighting device 210. The training data set may comprise previous operational state data; wherein the previous operational state data may comprise information indicative of durations of one or more previous operational states. The training data set may comprise the switching profile, the previous operational state data or both the switching profile and the previous operational state data.

For training of the untrained machine, the information contents in the training data set needs to be higher than a threshold. For example, the switching profile and/or the previous operational data may be unavailable in the case of a newly installed lighting device 210. In another example, the switching profile may comprise no switching but a continous On state or Off state. In these or similar examples, the training of the untrained machine may not be feasible. The controller 215, in such cases, may be then arranged for controlling the lighting device 210 according to predefined light settings. The predefined light settings may be user defined, or a maximum or a minimum value of the characteristic, e.g. the light device is always in an ON state with a certain color temperature.

The training data set may further comprise a second set of features, wherein the second set of features comprises one or more of: a weather condition, a measure of ambient light, an occupancy measure and an activity of the user. The training data set may comprise both the second set of features and the switching profile or the training data set may comprise only the second set of features. For example, an activity of the user may be correlated to the time of the day, such as in a factory or in a classroom, different activities are performed at specific time. Similarly, an occupancy measure may also be based on time of day. For outdoor settings, weather condition and/or a measure of ambient light may be useful data set for training the untrained machine. In such cases, the processing unit 213b may be further arranged for inferring, by the trained machine, the current time of day based on the second set of features or may be based on both the second set of features and the obtained operational data set.

The method 300 may be executed by computer program code of a computer program product when the computer program product is run on a processing unit of a computing device.

It should be noted that the above-mentioned embodiments illustrate rather than limit the invention, and that those skilled in the art will be able to design many alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between parentheses shall not be construed as limiting the claim. Use of the verb "comprise" and its conjugations does not exclude the presence of elements or steps other than those stated in a claim. The article "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The invention may be implemented by means of hardware comprising several distinct elements, and by means of a suitably programmed computer or processing unit. In the device claim enumerating several means, several of these means may be embodied by one and the same item of hardware. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage.

Aspects of the invention may be implemented in a computer program product, which may be a collection of computer program instructions stored on a computer readable storage device which may be executed by a computer. The instructions of the present invention may be in any interpretable or executable code mechanism, including but not limited to scripts, interpretable programs, dynamic link libraries (DLLs) or Java classes. The instructions can be provided as complete executable programs, partial executable programs, as modifications to existing programs (e.g. updates) or extensions for existing programs (e.g. plugins). Moreover, parts of the processing of the present invention may be distributed over multiple computers or processors or even the‘cloud’.

Storage media suitable for storing computer program instructions include all forms of nonvolatile memory, including but not limited to EPROM, EEPROM and flash memory devices, magnetic disks such as the internal and external hard disk drives, removable disks and CD-ROM disks. The computer program product may be distributed on such a storage medium, or may be offered for download through HTTP, FTP, email or through a server connected to a network such as the Internet.