Edge computing system with low power wide area network connectivity and autonomous or semi-autonomous machine learning
A mobile edge computing system, a communication network and a method of using a mobile edge computing device. The mobile edge computing system includes a communication module that establishes signal communication over numerous wireless communication protocols at least one of which uses a low power wide area network protocol. A logic device is made up of a distributed set of processing units including both a central processing unit and one or more processors configured for performing machine learning operations the latter of which include one or more of a graphical processor unit and a tensor processing unit. When the system receives event data that has been acquired by one or more of the communication module and a sensor, the system executes a trained machine learning model and conveys, using the low power wide area network protocol, an output that has been produced by the trained machine learning model.
1. A mobile edge computing system comprising:
a platform;
a communication module disposed on the platform and configured to establish signal communication over a plurality of wireless communication protocols at least one of which comprises a low power wide area network protocol such that the communication module transmits, over at least one of the wireless communication protocols, a signal that corresponds to an output that has been generated by a trained machine learning model;
at least one sensor;
a non-transitory computer-readable medium disposed on the platform and storing machine code thereon; and
a logic device disposed on the platform and comprising a distributed set of processing units comprising a central processing unit and at least one processor configured for performing machine learning operations and comprising at least one of a graphical processor unit and a tensor processing unit, wherein the mobile edge computing system, upon receipt of event data that has been acquired by at least one of the communication module and the at least one sensor, performs at least one of (i) event data preprocessing, (ii) feature extraction of the event data, (iii) segmentation of at least a portion of the event data that has undergone at least one of preprocessing and feature extraction into a training data set and a validation data set, (iv) utilization of at least one machine learning algorithm to provide an inference on at least a portion of the training data set and (v) validation of the inference with at least a portion of the validation data set to define the trained machine learning model that, upon receipt of at least one of (a) a portion of the event data that was not a part of at least one of the training data set and the validation data set and (b) subsequently acquired event data, either updates the trained machine learning model or executes the trained machine learning model.
2. The mobile edge computing system of claim 1 , wherein the logic device comprises a system-on-a-chip architecture.
3. The mobile edge computing system of claim 2 , wherein the system-on-a-chip architecture further comprises a microcontroller.
4. The mobile edge computing system of claim 3 , further comprising at least one machine learning library that cooperates with the microcontroller.
5. The mobile edge computing system of claim 1 , wherein the communication module defines a hybrid wireless communication module comprising:
a sub-module that is configured to transmit at least a portion of the event data using a short range protocol; and
a sub-module that is configured to transmit at least a portion of the event data using the at least one low power wide area network protocol.
6. The mobile edge computing system of claim 1 , wherein at least the platform, communication module, non-transitory computer-readable medium and logic device are arranged to define a wearable electronic device.
7. The mobile edge computing system of claim 6 , further comprising a plurality of devices each in signal communication with the wearable electronic device and configured as either a gateway or a bidirectional beacon.
8. The mobile edge computing system of claim 7 , wherein the plurality of devices comprise a plurality of gateways each signally cooperative with (i) one another to exchange at least a portion of the event data therebetween, (ii) the wearable electronic device using at least the low power wide area network protocol and (iii) an internet protocol network.
9. The mobile edge computing system of claim 7 , wherein the wearable electronic device and the plurality of gateways cooperate with one another to define a fog computing system.
10. The mobile edge computing system of claim 6 , wherein the non-transitory computer-readable medium, logic device and at least one machine learning algorithm cooperate to create the trained machine learning model with the event data.
11. The mobile edge computing system of claim 1 , wherein the non-transitory computer-readable medium, logic device and machine code cooperate to perform at least one of preprocessing and feature extraction of the event data prior to initiation of the communication.
12. The mobile edge computing system of claim 1 , wherein the non-transitory computer-readable medium, logic device and machine code cooperate to create the trained machine learning model with at least a portion of the event data.
13. The mobile edge computing system of claim 1 , wherein the logic device comprises a microcontroller.
14. A method of using a mobile edge computing device that has low power wide area network connectivity to operate a machine learning model, the method comprising:
configuring the mobile edge computing device to comprise:
a platform;
a communication module disposed on the platform and configured to transmit, over the low power side area network, a signal that corresponds to an output that has been generated by a trained machine learning model;
a non-transitory computer-readable medium disposed on the platform and storing machine code thereon; and
a logic device disposed on the platform and comprising a distributed set of processing units comprising:
a central processing unit; and
at least one processor configured for performing machine learning operations and comprising at least one of a graphical processor unit and a tensor processing unit;
upon the receiving event data into at least one of non-transitory computer-readable medium and logic device, having the mobile edge computing device, perform at least one of (i) event data preprocessing, (ii) feature extraction of the event data, (iii) segmentation of at least a portion of the event data that has undergone at least one of preprocessing and feature extraction into a training data set and a validation data set, (iv) utilization of at least one machine learning algorithm to provide an inference on at least a portion of the training data set and (v) validation of the inference with at least a portion of the validation data set to define the trained machine learning model that, upon receipt of at least one of (a) a portion of the event data that was not a part of at least one of the training data set and the validation data set and (b) subsequently acquired event data, either updates the trained machine learning model or executes the machine learning model that has been trained with at least a portion of the received event data; and
convey an output that has been produced by at least one of the trained and updated machine learning model to at least one wireless network using low power wide area network.
15. The method of claim 14 , wherein at least a portion of the event data is acquired by at least one of a plurality of sensors.
16. The method of claim 14 , wherein the mobile edge computing device further executes another communication comprising:
transmitting at least a portion of the received event data using the low power wide area network protocol prior to training the machine learning model; and
upon having the machine learning model be trained in a remote location, receiving the trained machine learning model using the low power wide area network protocol.
17. The method of claim 16 , wherein at least one of the communication and the another communication comprises using the mobile edge computing device and a plurality of gateways each in signal communication therewith to establish signal connection to an internet protocol network even upon failure of one of the plurality of gateways.
18. A mobile edge computing system comprising:
a platform;
a communication module disposed on the platform and configured to establish signal communication over a plurality of wireless communication protocols at least one of which comprises a low power wide area network protocol, the communication module further configured to transmit a signal that corresponds to an output that has been generated by a trained machine learning model;
at least one sensor;
a non-transitory computer-readable medium disposed on the platform and storing machine code thereon to define at least one machine learning algorithm as in-memory analytics to provide direct memory access thereof; and
at least one logic device disposed on the platform, wherein the mobile edge computing system, upon receipt of event data that has been acquired by at least one of the communication module and the at least one sensor, performs at least one of (i) event data preprocessing, (ii) feature extraction of the event data, (iii) segmentation of at least a portion of the event data that has undergone at least one of preprocessing and feature extraction into a training data set and a validation data set, (iv) utilization of at least one machine learning algorithm to provide an inference on at least a portion of the training data set and (v) validation of the inference with at least a portion of the validation data set to define the trained machine learning model that, upon receipt of at least one of (a) a portion of the event data that was not a part of at least one of the training data set and the validation data set and (b) subsequently acquired event data, either updates the trained machine learning model or executes the trained machine learning model.