Smart production system
Embodiments couple a corresponding IoT gateway to each IoT device, each IoT gateway monitoring for operation events of a smart contract of a distributed ledger, each IoT device and IoT gateway coupled to the distributed ledger. In response to a client initiating an operation of a first IoT device, embodiments generate a corresponding event by the smart contract and transmit an authorization request to an authorization system and in response receive an access token corresponding to the operation. Embodiments transmit the access token to one or more of the IoT gateways, each IoT gateway monitoring for the event and determining whether it corresponds to the first IoT device and then implementing the operation at the first IoT device. Embodiments determine a defect in any of the IoT devices using a trained Deep Convolutional Generative Adversarial Network (“DCGAN”) model coupled to the distributed ledger.
1. A method of monitoring a plurality of industrial Internet of Things (IoT) devices, using a distributed ledger, the method comprising:
coupling a corresponding IoT gateway to each IoT device of the plurality of IoT devices, each IoT gateway monitoring for operation events of a smart contract of the distributed ledger, each IoT device and IoT gateway coupled to the distributed ledger;
in response to a client initiating an operation of a first IoT device of the plurality of IoT devices, generating a corresponding operation event by the smart contract and transmitting an authorization request to an authorization system and receiving an access token corresponding to the operation, wherein the operation comprises performing a functionality of the first IoT device and the access token corresponds to an integrity measurement value for the first IoT device and an approval of the authorization request is based at least on the integrity measurement value;
transmitting the access token to one or more of the IoT gateways, each IoT gateway monitoring for the operation event and determining whether the operation event corresponds to the first IoT device and then implementing the operation at the first IoT device; and
determining a defect in any of the IoT devices using a trained Deep Convolutional Generative Adversarial Network (DCGAN) model coupled to the distributed ledger and comprising a corresponding discriminator and generator for each of the IoT devices, each of the IoT devices having a normal distribution set of data points, the DCGAN model monitoring transmitted data points from each of the IoT devices and detecting when the transmitted data points do not correspond to the normal distribution set of data points.
2. The method of claim 1 , further comprising training the DCGAN model, the training comprising:
each generator generating data points substantially close to the normal distribution set of data points for each corresponding IoT device, the data points including noise data;
each discriminator discriminating the generated data points from the normal distribution set of data points; and
repeating the generating and discriminating until each discriminator achieves an acceptable accuracy.
3. The method of claim 2 , wherein the discriminators comprise a convolutional neural network and the generators comprise a deconvolutional neural network.
4. The method of claim 1 , wherein the distributed ledger is a blockchain network.
5. The method of claim 1 , wherein the event comprises an event name, the event name comprising:
an extended validation code that encodes a desired operation;
a Uniform Resource Identifier (URI) that corresponds to a URI of the first IoT device; and
a dependency operation code for the desired operation.
6. The method of claim 5 , wherein each of the IoT gateways request operation events to be filtered by the extended validation code.
7. The method of claim 1 , the authorization request further transmitted to a cloud service provider that uses the integrity measurement value to determine whether to approve the request.
8. The method of claim 1 , wherein the token comprises an Ethereum Request for Comment-20 token.
9. The method of claim 1 , wherein the smart contract can determine a number of tokens held by the client, the number of tokens indicative of a role of the client and used to determine whether to authorize the operation.
10. A smart production system comprising:
a plurality of industrial Internet of Things (IoT) devices;
a corresponding IoT gateway coupled each IoT device of the plurality of IoT devices, each IoT gateway monitoring for operation events of a smart contract of a distributed ledger, each IoT device and IoT gateway coupled to the distributed ledger;
one or more processors, in response to a client initiating an operation of a first IoT device of the plurality of IoT devices, generating a corresponding operation event by the smart contract and transmitting an authorization request to an authorization system and receiving an access token corresponding to the operation, wherein the operation comprises performing a functionality of the first IoT device and the access token corresponds to an integrity measurement value for the first IoT device and an approval of the authorization request is based at least on the integrity measurement value;
the one or more processors transmitting the access token to one or more of the IoT gateways, each IoT gateway monitoring for the operation event and determining whether the operation event corresponds to the first IoT device and then implementing the operation at the first IoT device; and
a trained Deep Convolutional Generative Adversarial Network (DCGAN) model coupled to the distributed ledger and determining a defect in any of the IoT devices, the DCGAN model comprising a corresponding discriminator and generator for each of the IoT devices, each of the IoT devices having a normal distribution set of data points, the DCGAN model monitoring transmitted data points from each of the IoT devices and detecting when the transmitted data points do not correspond to the normal distribution set of data points.
11. The smart production system of claim 10 , the one or more processors further training the DCGAN model, the training comprising:
each generator generating data points substantially close to the normal distribution set of data points for each corresponding IoT device, the data points including noise data;
each discriminator discriminating the generated data points from the normal distribution set of data points; and
repeating the generating and discriminating until each discriminator achieves an acceptable accuracy.
12. The smart production system of claim 11 , wherein the discriminators comprise a convolutional neural network and the generators comprise a deconvolutional neural network.
13. The smart production system of claim 10 , wherein the distributed ledger is a blockchain network.
14. The smart production system of claim 10 , wherein the event comprises an event name, the event name comprising:
an extended validation code that encodes a desired operation;
a Uniform Resource Identifier (URI) that corresponds to a URI of the first IoT device; and
a dependency operation code for the desired operation.
15. The smart production system of claim 14 , wherein each of the IoT gateways request operation events to be filtered by the extended validation code.
16. The smart production system of claim 10 , the authorization request further transmitted to a cloud service provider that uses the integrity measurement value to determine whether to approve the request.
17. The smart production system of claim 10 , wherein the token comprises an Ethereum Request for Comment-20 token.
18. The smart production system of claim 10 , wherein the smart contract can determine a number of tokens held by the client, the number of tokens indicative of a role of the client and used to determine whether to authorize the operation.
19. A non-transitory computer-readable medium storing instructions which, when executed by at least one of a plurality of processors, cause the processors to monitor a plurality of industrial Internet of Things (IoT) devices, using a distributed ledger, the monitoring comprising:
coupling a corresponding IoT gateway to each IoT device of the plurality of IoT devices, each IoT gateway monitoring for operation events of a smart contract of the distributed ledger, each IoT device and IoT gateway coupled to the distributed ledger;
in response to a client initiating an operation of a first IoT device of the plurality of IoT devices, generating a corresponding operation event by the smart contract and transmitting an authorization request to an authorization system and receiving an access token corresponding to the operation, wherein the operation comprises performing a functionality of the first IoT device and the access token corresponds to an integrity measurement value for the first IoT device and an approval of the authorization request is based at least on the integrity measurement value;
transmitting the access token to one or more of the IoT gateways, each IoT gateway monitoring for the operation event and determining whether the operation event corresponds to the first IoT device and then implementing the operation at the first IoT device; and
determining a defect in any of the IoT devices using a trained Deep Convolutional Generative Adversarial Network (DCGAN) model coupled to the distributed ledger and comprising a corresponding discriminator and generator for each of the IoT devices, each of the IoT devices having a normal distribution set of data points, the DCGAN model monitoring transmitted data points from each of the IoT devices and detecting when the transmitted data points do not correspond to the normal distribution set of data points.
20. The computer-readable medium of claim 19 , further comprising training the DCGAN model, the training comprising:
each generator generating data points substantially close to the normal distribution set of data points for each corresponding IoT device, the data points including noise data;
each discriminator discriminating the generated data points from the normal distribution set of data points; and
repeating the generating and discriminating until each discriminator achieves an acceptable accuracy.