WO2020030585A1

WO2020030585A1 - Systems and methods using cross-modal sampling of sensor data in distributed computing networks

Info

Publication number: WO2020030585A1
Application number: PCT/EP2019/071016
Authority: WO
Inventors: Olaitan Philip OLALEYE; Abhishek MURTHY
Original assignee: Signify Holding B.V.
Priority date: 2018-08-07
Filing date: 2019-08-05
Publication date: 2020-02-13

Abstract

An edge node, connected computing system, and methods of data generation. The method includes collecting a first data signal related to a first sensor modality by a first sensor of one or more of edge nodes. A model is created from the first data signal. The model pertains to one or more data rate parameters affecting a second sensor modality of a second sensor of one or more of the edge nodes. The first and second sensor modalities are different. The one or more data rate parameters are set for the second sensor based on the model. A second data signal related to the second sensor modality is generated by the second sensor using the one or more data rate parameters.

Description

SYSTEMS AND METHODS USING CROSS-MODAL SAMPLING OF SENSOR DATA IN DISTRIBUTED COMPUTING NETWORKS

FIELD OF THE INVENTION

The present disclosure is directed generally to distributed computing networks having sensor-enabled edge nodes. BACKGROUND

There is an on-going trend in many industries to communicably connect together an increasingly large number of traditionally non-connected devices. Examples include home and building control systems (such as connected lighting systems), inventory tracking systems, the“internet of things”, and/or other“smart” or“connected” systems. Typically, the edge nodes of these distributed computing systems (e.g., the luminaires in a connected lighting system, such as the InterAct Office system under the Philips Lighting brand) include sensors and communication modules that enable the system as a whole to more effectively, efficiently, and/or automatically monitor and/or react to events in the relevant environment (e.g., home, office, warehouse, roadway, park, etc.). A typical characteristic of these systems is the limited availability of computation resources at the edge nodes (e.g., at the luminaires in a connected lighting system) and the low power bandwidth with which the nodes of the system are interconnected. As a result, the edge nodes are typically unable to send large amounts of data or perform complex computations in a timely manner.

Accordingly, there is a continued need in the art for systems and methods to facilitate more timely, accurate, and efficient generation and analysis of sensor data collected at resource-constrained edges nodes of distributed computing systems.

SUMMARY OF THE INVENTION

The present disclosure is directed to inventive systems and methods for reducing the sensor data generation at resource-constrained edge nodes of distributed computing networks. Various embodiments and implementations herein are directed to data generation systems in which the insights gained from one sensor modality are used to optimize data generation of other sensor modalities. The resulting amount of data produced by a distributed computing system can be reduced by generating data at rates required by each particular application to maintain sufficient accuracy. Consequently, the network load is reduced while accuracy is increased or maintained.

It is proposed that different sensor applications have different data needs. For example, motion detection sensors in an office environment only need to capture the activities of humans in the space, but the same style of sensors will need to capture a completely different set of dynamics in warehouses, which contain fork lifts and other vehicles. Other environmental conditions, such as temperature, humidity, etc., can also be very different. Additionally, variations are not taken into account throughout the lifecycle of a system, which may entail design, configuration, installation, commissioning, operation, data collection, analytics, and maintenance. As one example, people counting may only need low precision occupancy information, while higher-level activity recognition may need a very fine-grained occupancy signal from a motion detection sensor.

Despite these variations, many sensors generate data independent of the needs of the applications and the context in which they operate. For example, the same make and model of sensor can be bought (e.g., off the shelf) and installed for any number of different applications and then operate in the same manner for the entire life of the sensor.

Consequently, the edge nodes of connected systems, such as connected lighting systems, tend to generate very large amounts of data that is infeasible to analyze at the resource-constrained edge and/or transmit over the bandwidth-constrained wireless links.

All examples and features mentioned below can be combined in any technically possible way.

Generally, in one aspect, a method of generating data at edge nodes of a distributed computing network is provided. The method includes collecting a first data signal related to a first sensor modality by a first sensor of one or more of the edge nodes, creating a model from the first data signal, the model pertaining to one or more data rate parameters of a second sensor modality of a second sensor of one or more of the edge nodes, the first sensor modality being different than the first sensor modality, setting values of the one or more data rate parameters for the second sensor based on the model, and generating a second data signal by the second sensor using the one or more data rate parameters. In one embodiment, the first and second sensors are observing a same location. In one embodiment, first and second sensors are observing different locations.

According to an embodiment, the first data signal related to motion detection and the model is a mobility model defining movements of people or objects. In one embodiment, the first and second sensors are located with respect to different areas of a space, and mobility model includes rates of transitioning between the different areas, a proportion of occupants in each area, or a combination including at least one of the foregoing.

According to an embodiment, the first data signal relates to temperature and the model includes a temperature gradient map. In one embodiment, the one or more data rate parameters includes a sampling rate for the first sensor, a sensitivity of the first sensor, a precision of the first sensor, or a combination including at least one of the foregoing. In one embodiment, the generating includes collecting the second data signal. In one embodiment, the generating includes compressing and transmitting the second data signal. In one embodiment, the first data signal has a smaller data rate than that of the second data signal.

Generally, in another aspect, a node of a distributed computing network is provided. The node includes a controller configured to receive a first data signal related to a first sensor modality collected from a first sensor of one or more edge nodes of the distributed computing network; create a model from the first data signal, the model pertaining to one or more data rate parameters of a second sensor modality of a second sensor of one or more edge nodes of the distributed computing network, the first sensor modality being different than the second sensor modality; and set values of the one or more data rate parameters based on the model; and a communication module configured to transmit an instruction to generate a second data signal with the second sensor in accordance with the set values of the one or more parameters.

According to an embodiment, the node is a gateway or server for the distributed computing network. In one embodiment, the first and second sensors are observing a same location. In one embodiment, the one or more data rate parameters include a sampling rate, a sensor precision, a sensor sensitivity, or a combination including at least one of the foregoing.

Generally, in a further aspect, a connected computing system including a node as disclosed herein is provided in communication with a plurality of sensor-equipped edge nodes.

It should be appreciated that all combinations of the foregoing concepts and additional concepts discussed in greater detail below (provided such concepts are not mutually inconsistent) are contemplated as being part of the inventive subject matter disclosed herein. In particular, all combinations of claimed subject matter appearing at the end of this disclosure are contemplated as being part of the inventive subject matter disclosed herein. It should also be appreciated that terminology explicitly employed herein that also may appear in any disclosure incorporated by reference should be accorded a meaning most consistent with the particular concepts disclosed herein.

BRIEF DESCRIPTION OF THE DRAWINGS

In the drawings, like reference characters generally refer to the same parts throughout the different views. Also, the drawings are not necessarily to scale, emphasis instead generally being placed upon illustrating the principles of the invention.

FIG. 1 schematically illustrates a distributed computing network according to one embodiment disclosed herein.

FIG. 2 is a flowchart illustrating a method of reducing data generation of edge nodes of a distributed computing system.

FIG. 3 is a block diagram illustrating one example of setting data rate parameters for one sensor modality using data from another sensor modality.

FIG. 4 schematically illustrates a scenario in which two edge nodes are arranged to detect people or objects moving between different areas of a space, from which a mobility map can be created.

FIG. 5 illustrates a mobility model created with respect to the scenario depicted in FIG. 4.

DETAIFED DESCRIPTION OF EMBODIMENTS

The present disclosure describes various embodiments of distributed computing systems having sensor-enabled edge nodes. More generally, Applicant has recognized and appreciated that it would be beneficial to provide a system that utilizes the insight gained by one sensor modality to improve the data generated by one or more other sensor modalities. A particular goal of utilization of certain embodiments of the present disclosure is to reduce the amount of data generated by sensor-enabled edge nodes, particularly without unduly impacting the accuracy with which the data can be analyzed.

In view of the foregoing, various embodiments and implementations are directed to data generation systems in which the insights gained from one sensor modality are used to optimize data generation of other sensor modalities. The resulting amount of data produced by a distributed computing system can be reduced by generating data at rates required by each particular application to maintain sufficient accuracy. Consequently, the network load is reduced while accuracy is increased or maintained. It is proposed that different sensor applications have different data needs. For example, motion detection sensors in an office environment only need to capture the activities of humans in the space, but the same style of sensors will need to capture a completely different set of dynamics in warehouses, which contain fork lifts and other vehicles. Other environmental conditions, such as temperature, humidity, etc., can also be very different. Additionally, variations are not taken into account throughout the lifecycle of a system, which may entail design, configuration, installation, commissioning, operation, data collection, analytics, and maintenance. As one example, people counting may only need low precision occupancy information, while higher-level activity recognition may need a very fine-grained occupancy signal from a motion detection sensor.

Referring to FIG. 1, in one embodiment, a distributed computing network or system 10 having a plurality of edge nodes 12 is provided. It is noted that certain components discussed herein may be provided with alphabetic suffixes (e.g.,‘A’,‘B’) appended to a base reference numeral, such as the edge nodes 12A and 12B illustrated in FIG. 1. It is to be appreciated that any use of such alphabetic suffixes is intended to facilitate discussion with respect to particular instances of a component, while the base reference numeral without alphabetic suffix is intended to generally describe or refer to any and all such components.

In particular, the computing system 10 is arranged with two of the edge nodes 12 shown as luminaires (including the edge node 12 A, e.g., as part of a connected lighting system) and one as a surveillance camera (edge node 12B). It is to be appreciated that any other connected or“smart” computing system and corresponding maybe utilized, e.g., building control systems such as for automating heating, cooling, ventilation, security, noise suppression, etc., or other computing systems such as inventory tracking systems, the Internet of Things, etc.

At a minimum, each of the edge nodes 12 includes a sensor 14 (or multiple sensors) configured to collect data related to one or more parameters pertaining to the surrounding environment, a controller 15 configured to control operation of the respective edge node 12, and a communication interface or module 16 that enables data communication between the edge nodes 12 and/or with other nodes or devices. In one embodiment, the system 10 includes at least two sensors 14 of different modalities. For example, in the illustrated embodiment of FIG. 1, the edge node 12A may include the sensor 14A arranged as a passive infrared (PIR) or other motion detection sensor, while the edge node 12B may include the sensor 14B arranged as an electronic image sensor or any other sensor capable of capturing image data. Any other type or combination of sensor modalities may be used.

The term“centralized node” as used herein is intended to broadly refer to any designated network or processing equipment. Examples include a gateway 18, a server 20 (e.g., local network server), and a cloud computing implementation 22 (alternatively, the “cloud 22”). Thus, the centralized nodes generally provide additional computing resources for higher level processing needs, control operation of the system 10, enable certain features of the system 10, facilitate network traffic, etc. It is to be appreciated that the system 10 may include one or both of the server 20 and the cloud 22, which separately and/or together provide centralized computing resources for the system 10. For example, the central server 20 and/or the cloud 22 can be included to decompress and/or analyze the data collected by the sensors 14 and compressed at the edge nodes 12 as discussed in more detail herein.

It is to be appreciated that each of the edge nodes 12, the gateway 18, the server 20, and/or the cloud 22 may include suitable hardware and software to embody and enable the structure, features, and functionality disclosed herein. For example, each of the gateway 18, the server 20, and the cloud 22 may include sensors, controllers, and

communication modules akin to (e.g., but having more computational resources than) those described with respect to the edge nodes 12. Any such controllers may be, or include, a processor, memory, algorithm, or other hardware or software component.

Processors (e.g., for the controllers 15) may take any suitable form, such as a microcontroller, plural microcontrollers, circuitry, a single processor, or plural processors configured to execute software instructions. Similarly, memory (e.g., for the controllers 15) may take any suitable form or forms, including a volatile memory, such as random access memory (RAM), or non-volatile memory such as read only memory (ROM), flash memory, a hard disk drive (HDD), a solid state drive (SSD), or other data storage media. Memory may be used by a processor for the temporary storage of data during its operation. Data and software, such as the algorithms or software necessary to perform the methods and provide the features and functionality discussed herein, as well as an operating system, firmware, or other application, may be installed in memory. Communication modules as referred to herein (e.g., the communication modules 16) are arranged to enable communication between the components of the system 10. The communication module may be or include any module, device, or means capable of enabling the transmission and/or reception of a wired or wireless communication signal, e.g., a transmitter, receiver, transceiver, antenna, etc., utilizing technologies that include, but are not limited to Wi-Fi (e.g., IEEE 802.11), Bluetooth, cellular, Ethernet, Zigbee, etc. As discussed herein, of particular note are the more relatively low-power or resource constrained technologies, such as Zigbee.

The sensors 14 can be configured to collect any desired or selected data parameter related to the local environment, such as motion detection (e.g., radiated infrared energy), temperature, humidity, ambient noise level, ambient light level, etc. The system 10 may be arranged to utilize this data to automate or make more efficient certain features or functionality of its components, such as the edge nodes 12. For example, referring back to an embodiment in which the system 10 is a connected lighting system, the sensors 14 may be motion detection sensors that enable the system 10 to automatically turn lights on upon detection of movement and/or to turn off lights after a preset period of time elapses in which movement is not detected. Those of ordinary skill in the art will recognize other types of distributed computing systems and collected data that can be utilized to facilitate operation of these other systems.

It is also to be appreciated that the system 10 may be equipped with algorithms related to artificial intelligence, machine learning, artificial neural networks, etc. in order to enable advance decision making analytics and determination, such as context awareness, activity, event, or scene recognition, etc., based on the collected data. In one example, one or more algorithms may be utilized (e.g., by the server 20 and/or the cloud 22) to process the data collected by the sensors 14 to attempt to make a higher level determination or probabilistic guess as to a corresponding activity or scenario based on that data. For example, collected motion detection data may be analyzed to estimate a number of occupants in one or more designated areas. While motion detection data only indicates detected motion, the data can be analyzed for patterns that the algorithm leams, or is trained to, recognize or correlate to a corresponding event (or a likelihood of the corresponding event occurring), such as occupant count. For example, higher levels of detected motion in a concentrated area maybe interpreted as or correlated to a larger estimated occupant count. It is to be appreciated that this is merely one example and many other possibilities of higher level determinations made from collected sensor data will be appreciable to those of ordinary skill in the art. According to the systems and methods described herein, the information gained from a first sensor can be used to improve data generation at a second sensor. That is, optimizing the data collection of the sensors 14 to the conditions in the local environment will help optimize the overall data usage of the system 10. In one embodiment, the sensors may have different sensor modalities, such that information pertaining to the first modality of the first sensor is used to improve the ability of the second modality of the second sensor. In one embodiment, the first and second sensors are located in proximity to each other and/or arranged to observe the same events or phenomena. In one embodiment, the first and second sensors are spatially separated from each other (e.g., observing different rooms, areas, or locations), and the information pertaining to a first location of the first sensor is used to inform operation of the second sensor at a second location.

A method 30 describing a method of operating a distributed network (e.g., the system 10) is illustrated in FIG. 2. At step 32, a first data signal is collected by a first sensor. At step 34, a model is created from the first data. The model can be useful for translating the first data signal into a form that pertains to the second sensor. The model may define a relationship between the first sensor and the second sensor, such as by mapping certain values or patterns in the first data signal into one or more data rate parameters for the second sensor. At step 36, at least one data rate parameter for the second sensor is set in accordance with the model. At step 38, a second data signal is generated in accordance with the data rate parameters set in step 36. As used herein,“data rate parameter” is intended to mean any parameter related to altering an amount of data that is generated by a sensor, or that is desired to be transmitted between nodes of the system 10 (e.g., from the edge nodes 12 to the gateway 18, server 20, or cloud 22 for processing, which may include compression and reconstruction).

As one example, referring back to the embodiment of FIG. 1 , the surveillance camera (or other image sensor) 14B, such as included by the edge node 12B, may typically generate a significant amount of data, which can be difficult to accurately compress, transmit, and/or analyze at the edge of distributed networks. Accordingly, it may be desirable to instead analyze a comparatively low data rate signal, such as that of the motion detection sensor 14A of the edge node 12A, to see if any data rate parameters related to the sensor 14B (and/or the data generated thereby) can be implemented to reduce computational demands and/or increase analytical accuracy. As one example, if a motion detection sensor is not detecting any motion, then this can be utilized to decrease the quality and/or frequency of the images/video captured and/or transmitted by a nearby surveillance camera (e.g., as a lack of motion indicates there are no people or objects to actually track with the surveillance camera). As increased levels of motion is detected (e.g., indicating multiple people or objects, which may be more difficult for a surveillance camera to track at low quality settings), then the resolution, sampling frequency (frames per second), or other data rate parameters corresponding to the camera (or image data recorded by the surveillance camera) can be increased to ensure accuracy.

Parameters that impact the rate of generated data, such as the sensitivity or precision of the sensor (e.g., resolution of the saved image data saved by the image sensor 14B), sampling rate (e.g., frames per second of the video data captured by the image sensor 14B), etc. may be modified. By tuning these data rate parameters, the total amount of data required to be transmitted/analyzed can be reduced without negatively impacting accuracy and/or the accuracy of data analysis can be improved without increasing the amount of generated data.

A block diagram illustrating one example is shown in FIG. 3. In the scenario of FIG. 3, at least one of the edge nodes 12 includes a motion detection sensor (first sensor modality), while at least one other includes a thermopile or thermopile array (second sensor modality). The data generated from the edge nodes 12 can be transmitted to a centralized node, such as the gateway 18. Utilizing the two sensor modalities of this example, a mobility model (from the motion detection data) can be created at step 34A (i.e., a specific instance of the step 34). For example, mobility models can dictate the rate of movement of people or objects moving about a space (e.g., office, building, room, etc.), and also outline areas where mobility patterns are different from other spaces. It is noted that since thermopiles are included in this example, these can also be used to create a temperature model from collected temperature data, e.g., a temperature map indicating a temperature gradient across one or more areas.

The data model or models from step 34A can be used at steps 36A and/or 36B (two specific instances of the step 36) to set one or more data rate parameters. For ease of discussion herein,“data rate parameters 40” shall generally refer to any data rate parameters, including a sampling rate 40A and a precision 40B for the thermopiles, resulting respectively from the steps 36A and 36B. For example, in one embodiment the sampling rate 40A may take a value between about 2 Hz and 200Hz, although any other desired value can be selected. The precision 40B may take any desired resolution, such as 8x8 pixels, 16x16 pixels, 32x32 pixels, etc. The data rate parameters 40 can be derived using the highest frequency of motion conveyed by the mobility model created in the step 34A (e.g., selecting a faster sampling rate and/or higher resolution for higher frequency movements conveyed by the mobility model). In embodiments in which a temperature model is also or alternatively used, if the temperature gradient of the space is high, a higher resolution image can be generated, while if the temperature ranges are not very large, then a smaller resolution may be enough to capture all the dynamics.

An example for creating a mobility map can be appreciated in view of FIGS. 4-5. In this example, consider a room or space 42 with one exit and two sensor-equipped edge nodes, luminaires 12C and 12D, as depicted in FIG. 4. The interior of the room 42 can be divided into two parts, each corresponding to the sensor range of the sensor-equipped luminaires 12C and 12D, and an outside area. Accordingly, let T, h, and O denote the proportion of the occupants that are in the respective interior partitions and outside the room. The rates of transitioning among these partitions can be denoted as ai₂ (from T to I₂), u?i (from L· to Ii), aio, (from L· to O), aoi (from O to I2). It is noted that the proportions and the transfer rates can be estimated using the data from the sensors of the luminaires 12C and 12D, e.g., motion detection sensors. For example, the system include a machine learning algorithm, e.g., supervised or unsupervised, that is trained to map collected data values or patterns into the portions and transfer rates. Inverse modeling techniques, including optimization, may be used to update the parameters periodically.

Once we have these estimates, a model can be built, such as compartmental model 50 depicted in FIG. 5. The model describes how the proportions of occupants in the three regions evolve in time, as follows:

More exact rates can be determined by analyzing the motion of the objects and tracking the delay between crossings. For example, if an object is moving in such a way that it crosses luminaire 12C from left-to-right and then luminaire 12D from leflt-to -right, the delay between these two crossings will give a value for the rate 0112. In PIR or other motion detection sensors, the crossings can be detected using the analog responses using a pattern recognition algorithm such as a k-nearest neighbor algorithm. Additionally, dynamic time warping can be utilized as the distance metric to determine directionality while accounting for variances in movement speed. Advantageously, as described above, the model 50 can be utilized to facilitate data generation and analysis between the two luminaires 12C and 12D, or other edge nodes. For example, if the space 42 is an enclosed area, e.g., a room, with only a single exit, then the sensor of the luminaire 12C can collect samples at a relatively slow data rate and/or reduced sensitivity or precision until the model 50 indicates that people or objects are entering the space 42 (e.g., as detected by the sensor of the luminaire 12D), since any person or object would need to move past, and be detected by, the luminaire 12D before reaching the luminaire 12C. Similarly, if the rates ai₂ and aio indicate that several people, objects, or other entities are moving out of the space 42, this information can be transmitted and utilized to cause a third edge node (not illustrated) located outside of the space 42 to begin sampling at a higher frequency and/or with a greater sensitivity or precision to capture a sufficient level of data to more accurately track, monitor, or observe all of the people or objects leaving the space 42.

While several inventive embodiments have been described and illustrated herein, those of ordinary skill in the art will readily envision a variety of other means and/or structures for performing the function and/or obtaining the results and/or one or more of the advantages described herein, and each of such variations and/or modifications is deemed to be within the scope of the inventive embodiments described herein. More generally, those skilled in the art will readily appreciate that all parameters, dimensions, materials, and configurations described herein are meant to be exemplary and that the actual parameters, dimensions, materials, and/or configurations will depend upon the specific application or applications for which the inventive teachings is/are used. Those skilled in the art will recognize, or be able to ascertain using no more than routine experimentation, many equivalents to the specific inventive embodiments described herein. It is, therefore, to be understood that the foregoing embodiments are presented by way of example only and that, within the scope of the appended claims and equivalents thereto, inventive embodiments may be practiced otherwise than as specifically described and claimed. Inventive embodiments of the present disclosure are directed to each individual feature, system, article, material, kit, and/or method described herein. In addition, any combination of two or more such features, systems, articles, materials, kits, and/or methods, if such features, systems, articles, materials, kits, and/or methods are not mutually inconsistent, is included within the inventive scope of the present disclosure.

The phrase“and/or,” as used herein in the specification and in the claims, should be understood to mean“either or both” of the elements so conjoined, i.e., elements that are conjunctively present in some cases and disjunctively present in other cases. Multiple elements listed with“and/or” should be construed in the same fashion, i.e.,“one or more” of the elements so conjoined. Other elements may optionally be present other than the elements specifically identified by the“and/or” clause, whether related or unrelated to those elements specifically identified. As used herein in the specification and in the claims,“or” should be understood to have the same meaning as“and/or” as defined above.

It should also be understood that, unless clearly indicated to the contrary, in any methods claimed herein that include more than one step or act, the order of the steps or acts of the method is not necessarily limited to the order in which the steps or acts of the method are recited.

Claims

CLAIMS:

1. A method (30) of generating data at edge nodes of a distributed computing network (10), comprising:

collecting a first data signal (32) related to a first sensor modality by a first sensor of one or more of the edge nodes;

creating a model (34) based on the first data signal, the model pertaining to one or more data rate parameters of a second sensor modality of a second sensor of one or more of the edge nodes, the one or more data rate parameters relating to an amount of data generated by the second sensor modality, wherein the second sensor modality being different than the first sensor modality;

adjusting values of the one or more data rate parameters (36) for the second sensor based on the model; and

generating a second data signal by the second sensor using the one or more data rate parameters (38).

2. The method of claim 1, wherein the first and second sensors are observing a same location.

3. The method of claim 1, wherein the first and second sensors are observing different locations.

4. The method of claim 1, wherein the first data signal related to motion detection and the model is a mobility model defining movements of people or objects.

5. The method of claim 4, wherein the first and second sensors are located with respect to different areas of a space, and mobility model includes rates of transitioning between the different areas, a proportion of occupants in each area, or a combination including at least one of the foregoing.

6. The method of claim 1, wherein the first data signal relates to temperature and the model includes a temperature gradient map.

7. The method of claim 1, wherein the one or more data rate parameters includes a sampling rate for the first sensor, a sensitivity of the first sensor, a precision of the first sensor, or a combination including at least one of the foregoing.

8. The method of claim 1, wherein the generating includes collecting the second data signal.

9. The method of claim 1, wherein the generating includes compressing and transmitting the second data signal.

10. The method of claim 1, wherein the first data signal has a smaller data rate than that of the second data signal.

11. A node of a distributed computing network, comprising:

a controller configured to:

receive a first data signal (32) related to a first sensor modality collected from a first sensor of one or more edge nodes of the distributed computing network;

create a model (34) based on the first data signal, the model pertaining to one or more data rate parameters of a second sensor modality of a second sensor of one or more edge nodes of the distributed computing network, the one or more data rate parameters relating to an amount of data generated by the second sensor modality, wherein the first sensor modality being different than the second sensor modality; and

adjust values of the one or more data rate parameters (36) based on the model; and

a communication module configured to transmit an instruction to generate a second data signal (38) with the second sensor in accordance with the set values of the one or more parameters.

12. The node of claim 11, wherein the node is a gateway or server for the distributed computing network.

13. The node of claim 11, wherein the first and second sensors are observing a same location.

14. The node of claim 11, wherein the one or more data rate parameters include a sampling rate, a sensor precision, a sensor sensitivity, or a combination including at least one of the foregoing.

15. A connected computing system including the node of claim 11 in

communication with a plurality of sensor-equipped edge nodes.