FLEXIBLE SENSOR NETWORK IN AN IOT ENVIRONMENT

Abstract

The present invention relates to a sensor network in an IoT communication environment, wherein the sensor network comprises: • a server configured for communication with at least one sensor unit via a bidirectional communication link; • at least one first sensor unit for detecting physical data, wherein the sensor unit comprises: ∘ a communication interface configured for bidirectional communication via the bidirectional communication link with a second sensor unit and / or the server, ∘ at least one first sensor for detecting physical data, ∘ a processor for controlling the operation of the first sensor unit; wherein the processor has a programming interface for changing the operation of the sensor unit.

Description

[0001] The present invention relates to the technical field of a sensor network and a method for operating the sensor network, in particular the present invention relates to a sensor network in an IoT environment.

[0002] Various service providers, including network operators like Deutsche Telekom, are installing sensors, particularly IoT sensors, in urban areas as part of various IoT projects. These sensors are usually permanently installed and are designed to provide information about predefined physical environmental parameters. For example, sensors can be installed that are specifically designed to determine parking space occupancy, measure air quality (e.g., particulate matter), detect moving objects on a road, or measure traffic flow at a traffic light. To capture many physical parameters of an urban area, a significant number of such sensors are required, all of which must be installed, interconnected, and operated in a complex process.Even with modern, cost-effective network connections, such sensor networks are therefore usually very complex and, for this reason, very expensive and partially not economically viable for service providers.

[0003] Therefore, common practice is to install and operate a variety of sensors to determine urban parameters, specifically for individual applications and special tasks.

[0004] The object of the invention is therefore to provide a sensor network and a corresponding method that enables more efficient, in particular more cost-effective, and more flexible operation of the sensor network.

[0005] This task is solved using the characteristics of independent claims.

[0006] According to the invention, a sensor network in an IoT communication environment is specified, wherein the sensor network comprises: a server configured for communication with at least one sensor unit via a bidirectional communication link; at least one first sensor unit for detecting physical data, wherein the sensor unit comprises: ∘ a communication interface configured for bidirectional communication via the bidirectional communication link with a second sensor unit, the server and / or a forwarding unit, ∘ at least one first sensor for detecting physical data, ∘ a processor for controlling the operation of the first sensor unit; wherein the processor has a programming interface for changing the operation of the sensor unit.The core idea of ​​the invention is that the bidirectional communication link between the sensor unit and the server not only allows the sensor unit to transmit the measured physical parameters to the server, but also enables the server to communicate with the sensor unit, i.e., to send "commands" via the programming interface. Combined with the ability of the sensor unit(s) to change their operating mode via the programming interface, this allows a single sensor unit or detector head to be used for different application scenarios by "reprogramming" the sensor unit – meaning the same hardware can be used for different measurements. As a concrete example, various physical parameters are measured using a single spectrometer as the sensor unit.If necessary, this sensor unit must not only be reprogrammed but also adjusted to the new measurement in terms of hardware – for example, by briefly heating it and / or changing the applied voltages and / or currents. Such flexible sensors can be based on a gallium nitride (GaN) / aluminum gallium nitride (AIGaN) quantum sensor design. This makes a sensor unit significantly more flexible and thus considerably reduces costs for a service provider, as the total number of sensors required to measure various urban parameters is significantly reduced.

[0007] Preferably, modern network technologies such as NB-IoT, CatM, and / or 5G are used to establish the bidirectional communication link because they can be operated with very low power consumption, which significantly increases the operating time of the sensor units. A further advantage of the sensor network's functionality, adapted to current needs, is the increased value it brings to the service provider's mobile network.

[0008] By providing a communication interface that can optionally establish a bidirectional connection with a second sensor unit, a server, or a relay unit such as a router, the sensor unit can create a mesh network. In this network, the individual sensor units only need to transmit over short distances, significantly reducing their power consumption. For example, a router could be conveniently located to serve a sensor network comprising multiple sensor units. The individual sensor units then transmit their measured data to the router, which in turn sends the data to the server either wirelessly or via a wired connection. This offers the advantage of making it much easier to replace batteries in individual relay units or connect them to a power supply.

[0009] The sensor is therefore configured to record various physical parameters or data and / or to measure a given physical parameter in a modified manner. For example, if the sensor unit was initially intended to measure air quality once a day and it turns out that this time interval was set too long, the solution according to the invention makes it possible to reprogram the sensor unit to measure air quality, for example, once an hour.In particular, a detector head can be provided on the sensor for measuring various physical parameters. This detector head can incorporate multiple sensor components, and / or an existing sensor component can be reprogrammed. The use of different sensor components can then be leveraged to detect a wide range of physical parameters that are not accessible to a single reprogrammable sensor component. For example, a first sensor component could be a motion detector, while a second sensor component could be an air quality detector. This advantageously allows the sensor units to be used in a broader range of applications, rather than being limited to a single, predefined use case.

[0010] Preferably, the server is configured to transfer program code to the programming interface via the bidirectional communication link.

[0011] This offers the advantage that the service provider can centrally modify the functionality of individual sensor units via the server. Service providers have the option of modifying the functionality of individual sensor units or all sensor units that make up the sensor network. To enable the targeted transmission of program code to individual sensor units, a database containing unique identification information for each sensor unit is stored on the server. By modifying the functionality of the sensor units using program code, the sensor units are reprogrammed "permanently"—until a new program code is sent again. This is significantly more efficient than if the server were to change the functionality of the sensor units by transmitting a new control command to each sensor unit on a case-by-case basis.

[0012] Ideally, the server is designed to manage and analyze sensor data from at least a plurality of sensor units.

[0013] This offers the advantage of reducing the power consumption of the sensor units by offloading the computing power required for analyzing the sensor data from the server. Another advantage is that the analysis of the collected data provides a more comprehensive overview of the urban situation under investigation, such as parking space occupancy. Furthermore, the increased data volume also reduces statistical uncertainties.

[0014] Preferably, the server runs artificial intelligence processes that are trained to dynamically adapt the sensor programming to the task at hand. Methods such as deep learning and reinforcement learning can be used for this purpose.

[0015] This offers the advantage of automatically modifying the functionality of one or more sensor units, with program code generated based on the analysis of a single or multiple sensor units. For example, the server might analyze data collection frequency and determine that it is insufficient to make accurate assessments of an urban environment. The server can then generate program code to increase the data acquisition frequency of the sensor units. Alternatively, the server could generate program code to optimize the power consumption of the sensor units. Furthermore, it is possible to add or remove other sensor components. This allows, for instance, the integration of an air pollution detector alongside a motion detector.

[0016] Within the scope of the invention, it is also possible for a programmer to create a program code and transmit it to the sensor units via the communication link. This allows the functionality of a sensor unit to be specifically modified in a highly individualized way, which is not included in the functionality of the artificial intelligence. The respective program codes can have a "tag" that specifies whether the program code may be modified, may not be modified, or by whom it may be modified.

[0017] In one embodiment, the server is configured to send control commands to external actuators connected to the communication network, based on the sensor data transmitted to it.

[0018] The communication network can be, in particular, the internet. For example, signaling devices that respond to control commands from the server can be connected to this communication network. Based on the measurement data transmitted by the sensor units, the server can, for instance, determine how best to control one or more traffic lights to ensure an unimpeded flow of traffic. By evaluating measurements from an air pollution detector, it can be signaled that access to certain areas should be avoided or prohibited when pollution levels are high.

[0019] Ideally, bidirectional communication links include wireless radio links and / or wired links.

[0020] This offers the advantage that when installing sensor units in the sensor network, a complex wired connection is not required for each sensor unit, or even for any sensor unit at all. The individual sensor units can communicate with each other and, if necessary, with the server via wireless radio links, thereby establishing a so-called mesh network. However, modern communication standards such as 5G also allow for the efficient operation of a sensor network without using a mesh network. A specially selected sensor unit within the sensor network, or a dedicated relay unit such as a router, can forward the individual sensor data transmitted wirelessly from the respective sensor units to the communication network, either via cable or wirelessly, thus enabling its transmission to the server.

[0021] This offers the advantage that all sensor units, except for the specifically selected sensor unit, only need to transmit their collected sensor data over a short distance, thus reducing their power consumption. The specifically selected sensor unit or the relay unit can be connected to a dedicated power supply if necessary. This arrangement significantly reduces the installation effort for the sensor network.

[0022] According to an advantageous embodiment, the wireless radio connections are based on an energy-saving standard.

[0023] This offers the advantage of further reducing the energy consumption of the sensor units. The low-energy standard can be NB-lot, CatM, 5G, and / or LPWAN. A Low Power Wide Area Network (LPWAN) describes a class of network protocols for connecting low-energy devices, such as battery-powered sensors, to a network server. The protocol is designed to achieve a long range and low energy consumption of the end devices at low operating costs.

[0024] The sensor units may conveniently include a sensor element adapter for a sensor head.

[0025] This offers the advantage that various sensor components can be easily provided to the sensor unit if a single reprogrammable sensor component is insufficient. For example, if a sensor unit only has a motion detector in its sensor head, this sensor head can be removed and replaced with another sensor head, which is inserted into the sensor unit using the sensor element adapter. The other sensor head could, for example, include an air pollution detector. The operation of the sensor unit is adapted to the air pollution detector as described above by uploading new program code to the sensor unit via the programming interface, which modifies the sensor unit's operation to match the air pollution detector.As explained above, by transferring new program code to the sensor unit via the programming interface, a "single" sensor component can also be specifically reprogrammed to measure other physical parameters.

[0026] In principle, the sensor unit can be powered by a dedicated power supply or a battery. However, since IoT sensors are flexible, numerous, and often installed in many locations, a dedicated power supply is usually impractical, making battery power the preferred option.

[0027] However, such a battery unit has a limited lifespan, so an energy generation unit may be incorporated into the sensor unit in addition to the battery. For example, if the sensor unit is located outdoors, a solar panel can be used as the energy generation unit to charge the battery. If the sensor unit is located under a road, a piezoelectric element can be used as the energy generation unit, generating energy through pressure changes caused, for example, by cars on the road.

[0028] According to a further aspect of the invention, a method for operating a sensor network described above in an IoT communication environment is specified, wherein the server has program code for changing the functionality of the sensor unit by means of the bidirectional communication link of the programming interface of the sensor unit.

[0029] The advantages of this method essentially correspond to the advantages described above in connection with the sensor network.

[0030] Preferred embodiments of the present invention are explained below with reference to the accompanying figure: Fig. 1: shows an efficient and flexible sensor network in an IoT environment.

[0031] Numerous features of the present invention are explained in detail below with reference to preferred embodiments. The present disclosure is not limited to the specific combinations of features mentioned. Rather, the features mentioned here can be combined arbitrarily to form embodiments according to the invention, unless expressly excluded below.

[0032] Fig. 1 demonstrates an efficient and flexible sensor network 100 in an IoT environment.

[0033] The sensor network 100 comprises a plurality of sensor units 110, a relay unit 150, and a server 160, wherein the sensor units 110, the relay unit 150, and the server 160 are interconnected via a bidirectional connection 170 and can exchange data, together forming a communication network 105. The bidirectional connection 170 can be a radio connection or a wired connection. Preferably, the sensor units 110 are interconnected with each other and with the relay unit 150 via a radio connection 170a, whereas the relay unit 150 is connected to the server 160 via a wired connection 170b. The relay unit 150 can also be a base station that enables access to a communication network, in particular the Internet, of a network provider.

[0034] One of the sensor units 110a is shown in detail. The sensor units 110 have a communication interface 112 through which data can be exchanged with the communication network 105. A processor 114 of the sensor unit 110 can forward data measured by its sensor 116 to the server 160 via the communication interface 112. If the operating principle of one of the sensor units 110 is to be changed, program code can be sent from the server 110 to the communication interface 112 of the sensor unit 110 via the bidirectional connections 170. This program code is forwarded from the communication interface 112 to a programming interface 118, which automatically implements the program code on the processor 114 of the sensor unit. After the implementation of the new program code, the sensor unit can operate with a modified function. The sensor unit also has a battery unit 120.

[0035] The corresponding program code can be generated by an artificial intelligence implemented on server 160. For example, the artificial intelligence on server 160 can analyze the collected sensor data from sensor unit 110 and determine that measurement rates need to be changed and / or other sensor components should be activated.

[0036] This will be illustrated by an example: Sensor unit 110 contains at least two different sensor components, one designed to measure physical parameters of a first kind and the other designed to measure physical parameters of a second kind. For clarity, and without limiting the scope of this example, it is assumed that the first sensor component is a motion detector and the second sensor component is an air pollution detector.

[0037] It is often undesirable for the sensor unit 110, despite containing both sensor components, to measure and transmit the corresponding data at a high frequency, as this would reduce the sensor unit's battery life. However, communication with the server 160 now makes it possible to react adaptively to changing environmental conditions and adjust the measurements of the sensor units 110 accordingly.

[0038] Assuming that the sensor unit 110a is primarily intended to detect movements and transmit corresponding feedback to a traffic light to control traffic flow, but that the sensor unit 110a is also technically capable of measuring air pollution.

[0039] A phenomenon frequently observed, especially in cities, is the need to monitor air pollution. It is known that air pollution can reach high levels, particularly on busy roads. These roads are equipped with sensor units (110) designed to measure air pollution. In this context, it is often interesting to observe how air pollution spreads across the urban area.

[0040] However, it would often be an economically disproportionately high expense to install specially designed sensor units 110 throughout the city, which are specifically intended to record air pollution levels with a high frequency of data acquisition.

[0041] For example, server 160 can receive information about current wind direction and speed from appropriately trained sensor units 110. Server 160's artificial intelligence can analyze this information and calculate the direction in which the fine dust particles generated by exhaust fumes are blown. If the server 160 has stored the local positions of the sensor units 110a, its artificial intelligence can selectively send program code via communication network 105 to the sensor units 110a located in the direction of the wind. This code modifies the sensor units 110a's function, enabling them to collect data on air pollution as an alternative or additional function to motion detection.

[0042] Similarly, it is possible to adaptively adjust measurement rates to the environment. Based on the transmitted data, the artificial intelligence on server 160 can estimate whether increasing or decreasing the measurement rates is advisable at a particular time and generate program code that controls the sensor units 110 accordingly. For example, traffic flow can be recorded at a lower frequency at midnight than during rush hour.

[0043] Artificial intelligence can also control which type of measurement should be prioritized.

[0044] Another example: in the city center, ground sensors 110 are being installed to sequentially perform various measurement tasks. These ground sensors 110 are designed for traffic measurement (pedestrians, cyclists, cars, trucks), air quality measurement (particulate matter, NOx, SOx), and measurement of organic components in the air (pollen, viruses, etc.). The sensor network 100 can now be adapted so that pollen measurements are suppressed during periods of high traffic (for example, rush hour), as the results cannot be meaningfully evaluated during rush hour due to the high levels of pollutants. In return, the accuracy of the traffic measurement is increased. At other times, pollen levels can be determined more accurately, for example, in spring or summer, or when there is little traffic.

Claims

1. Sensor network in an IoT communication environment, wherein the sensor network (100) comprises: • a server (160) configured to communicate with at least one sensor unit (110) via a bidirectional communication link (170); • at least one first sensor unit (110) for detecting physical data, comprising: ∘ a communication interface (112) configured for bidirectional communication via the bidirectional communication link (170) with a second sensor unit and / or the server (160), ∘ at least one first sensor (116) for detecting physical data, ∘ a processor (114) for controlling the operation of the first sensor unit (110), characterized by that The processor (114) has a programming interface (118) for changing the operation of the sensor unit (110).

2. Sensor network according to claim 1, characterized by the fact thatthe server (160) is designed to transmit program code to the programming interface (118) via the bidirectional communication link (170).

3. Sensor network according to one of the preceding claims, characterized by the fact that the server (160) is trained to manage and / or analyze the sensor data from at least a plurality of sensor units (110).

4. Sensor network according to claim 3, characterized by the fact that An artificial intelligence is implemented on the server (160) that is trained to create the program code.

5. Sensor network according to claim 4, characterized by the fact that the artificial intelligence is trained to create the program code based on an evaluation of the sensor data from the majority of sensor units (110).

6. Sensor network according to one of the preceding claims, characterized by the fact thatthe server (160) is configured to send control commands to external actuators, which are connected to a communication network (105), based on the sensor data transmitted to it, wherein the server (160) is also connected to the communication network (105), in particular the Internet.

7. Sensor network according to one of the preceding claims, characterized by the fact that the bidirectional communication link (170) includes wireless radio links and / or wired links.

8. Sensor network according to claim 7, characterized by , the wireless radio connections are based on an energy-saving standard.

9. Sensor network according to any of the preceding claims, characterized by the fact that the first sensor unit (110) has a sensor element adapter.

10. Sensor network according to one of the preceding claims, characterized by the fact that the first sensor unit has a power generation unit.

11. Method for operating a sensor network according to one of claims 1-10 in an IoT communication environment, characterized by that The server (160) transmits program code to change the functionality of the sensor unit (110) via the bidirectional communication link of the programming interface (118) of the sensor unit (110).