Apparatus and method for monitoring and controlling of a neural network using another neural network implemented on one or more solid-state chips
A method of operating an apparatus using a control system that includes at least one neural network. The method includes receiving an input value captured by the apparatus, processing the input value using the at least one neural network of the control system implemented on first one or more solid-state chips, and obtaining an output from the at least one neural network resulting from processing the input value. The method may also include processing the output with another neural network implemented on solid-state chips to determine whether the output breaches a predetermined condition that is unchangeable after an initial installation onto the control system. The aforementioned another neural network is prevented from being retrained. The method may also include the step of using the output from the at least one neural network to control the apparatus unless the output breaches the predetermined condition. Similar corresponding apparatuses are described.
1. A method of operating an apparatus using a control system that includes at least one neural network, comprising the steps of:
receiving input data captured by the apparatus;
processing the input data using the at least one neural network of the control system, the at least one neural network having at least one hidden layer;
obtaining an output from the at least one neural network resulting from processing the input data; and
receiving, using a second neural network having at least one hidden layer, the output from the at least one neural network;
generating, using the second neural network, a confidence level of the output; and
controlling the apparatus using the output if the confidence level is not lower than the predetermined level.
2. The method of claim 1 , wherein the at least one neural network and the second neural network are implemented on physically different machines.
3. The method of claim 2 , wherein the different machines physically comprise autonomous machines.
4. The method of claim 1 , wherein the at least one neural network and the second neural network are instantiated on different virtual machines on the apparatus.
5. The method of claim 1 , wherein the at least one neural network and the second neural network are deployed in different memory spaces.
6. The method of claim 1 , wherein the different memory spaces are on physically different machines.
7. The method of claim 1 , further comprising training the at least one neural network, wherein training the at least one first neural network comprises training the neural network using globally collected training data including training data from numerous autonomous machines and training data from other sources.
8. The method of claim 1 , further comprising re-training the at least one neural network, wherein re-training the at least one neural network comprises training the at least one neural network using locally collected training data from at least one autonomous machine.
9. The method of claim 1 , further comprising re-training the at least one neural network, wherein re-training the at least one neural network comprises training the at least one neural network using globally collected training data and locally collected training data from at least one autonomous machine.
10. The method of claim 1 , further comprising re-training the at least one neural network, wherein re-training the at least one neural network comprises training the at least one neural network using locally collected training data from at least one autonomous machine.
11. The method of claim 1 , further comprising re-training the at least one neural network if the confidence level of the output is lower than the predetermined level.
12. A system for controlling a plurality of machines, comprising:
a controller;
an apparatus in communication with the controller and being operated in part by the controller, the apparatus comprising:
an input device constructed to generate input data, and
at least one neural network constructed to receive the input data and to generate an output, the at least one neural network having at least one hidden layer; and
a second neural network with at least one hidden layer constructed to use the output from the at least one neural network to generate a confidence level of the output,
wherein the controller is constructed to operate the apparatus using the output unless the confidence level of the output is below a predetermined level.
13. The system of claim 10 , wherein the at least one neural network and the second neural network are implemented on physically different machines.
14. The system of claim 13 , wherein the physically different machines comprise autonomous machines.
15. The system of claim 10 , wherein the at least one neural network and the second neural network are instantiated on different virtual machines.
16. The system of claim 10 , wherein the at least one neural network and the second neural network are instantiated in different memory spaces on the apparatus.
17. The system of claim 10 , wherein the at least one neural network is trained using globally collected training data including training data from numerous autonomous machines and training data from other sources.
18. The system of claim 10 , wherein the at least one neural network was trained using locally collected training data from at least one autonomous machine.
19. The system of claim 10 , wherein the at least one neural network is trained using globally collected training data and locally collected training data from at least one autonomous machine.
20. The system of claim 10 , wherein the at least one neural network was retrained using locally collected training data from at least one autonomous machine.