Neural network-based system for flight condition analysis and communication
A pilot support system includes: a processing circuit; and memory storing instructions that, when executed by the processing circuit, cause the processing circuit to: receive input data regarding a current state of a vehicle; encode the input data to generate encoded input data; supply the encoded input data to a trained statistical model; compute a current context of the vehicle based on the input data using the trained statistical model; compute one or more pilot feedback indicators based on the current context using the trained statistical model; and provide the one or more pilot feedback indicators to a cockpit of the vehicle.
1. A pilot support system comprising:
a processing circuit; and
memory storing instructions that, when executed by the processing circuit, cause the processing circuit to:
receive input data regarding a current state of a vehicle;
encode the input data to generate encoded input data comprising a plurality of vectors of summary values computed from the input data, comprising:
binning the input data into time windows, each time window including a plurality of data samples of the input data; and
computing the plurality of vectors of summary values, each vector of the plurality of vectors of summary values being computed based on the plurality of data samples binned into a corresponding one of the time windows;
supply the plurality of vectors of summary values of the encoded input data to a trained statistical model configured to cluster data corresponding to a current operational state of the vehicle into different contexts of the vehicle;
compute a current context of the vehicle based on the plurality of vectors of summary values of the encoded input data using the trained statistical model;
compute one or more pilot feedback indicators based on the current context using the trained statistical model; and
provide the one or more pilot feedback indicators to a cockpit of the vehicle.
2. The pilot support system of claim 1 , wherein the trained statistical model comprises a neural network.
3. The pilot support system of claim 2 , wherein the neural network comprises a Restricted Boltzmann Machine.
4. The pilot support system of claim 1 , wherein the input data comprises cockpit sensor data and cockpit controls data.
5. The pilot support system of claim 4 , wherein the vehicle is an aircraft,
wherein the cockpit sensor data comprises: aircraft pitch and aircraft yaw, and
wherein the cockpit controls data comprises: flight stick position.
6. The pilot support system of claim 1 , wherein the trained statistical model is trained using training data corresponding to operation of a vehicle by a pilot, the training data comprising:
training sensor data;
training cockpit control data; and
semantic labels associated with different contexts.
7. The pilot support system of claim 1 , wherein
the vector of summary values comprises:
a mean;
a maximum; and
a minimum.
8. The pilot support system of claim 1 , wherein the pilot feedback indicators comprise instruments relevant to the current context, and
wherein the instructions to provide the one or more pilot feedback indicators comprise instructions that, when executed by the processing circuit, cause the processing circuit to illuminate lights associated with the instruments relevant to the current context.
9. The pilot support system of claim 1 , wherein the pilot feedback indicators comprise proposed control actuations, and
wherein the instructions to provide the one or more pilot feedback indicators comprise instructions that, when executed by the processing circuit, cause the processing circuit to display a proposed direction for a cockpit control to be actuated.
10. The pilot support system of claim 1 , wherein the pilot feedback indicators comprise a semantic label, and
wherein the instructions to provide the one or more pilot feedback indicators comprise instructions that, when executed by the processing circuit, cause the processing circuit to display the semantic label on a display device in the cockpit.
11. A method for providing feedback to a pilot, the method comprising:
receiving, by a processing circuit, input data regarding a current state of a vehicle operated by the pilot;
encoding, by the processing circuit, the input data to generate encoded input data comprising a plurality of vectors of summary values computed from the input data, the generating the encoded input data comprising:
binning the input data into time windows, each time window including a plurality of data samples of the input data; and
computing the plurality of vectors of summary values, each vector of the plurality of vectors of summary values being computed based on the plurality of data samples binned into a corresponding one of the time windows;
supplying, by the processing circuit, the plurality of vectors of summary values of the encoded input data to a trained statistical model configured to cluster data corresponding to a current operational state of the vehicle into different contexts of the vehicle;
computing, by the processing circuit, a current context of the vehicle based on the plurality of vectors of summary values of the encoded input data using the trained statistical model;
computing, by the processing circuit, one or more pilot feedback indicators based on the current context using the trained statistical model; and
providing, by the processing circuit, the one or more pilot feedback indicators to a cockpit of the vehicle.
12. The method of claim 11 , wherein the trained statistical model comprises a neural network.
13. The method of claim 12 , wherein the neural network comprises a Restricted Boltzmann Machine.
14. The method of claim 11 , wherein the input data comprises cockpit sensor data and cockpit controls data.
15. The method of claim 14 , wherein the vehicle is an aircraft,
wherein the cockpit sensor data comprises: aircraft pitch and aircraft yaw, and
wherein the cockpit controls data comprises: flight stick position.
16. The method of claim 11 , wherein the trained statistical model is trained using training data corresponding to operation of a vehicle by a pilot, the training data comprising:
training sensor data;
training cockpit control data; and
semantic labels associated with different contexts.
17. The method of claim 11 , wherein
the vector of summary values comprises:
a mean;
a maximum; and
a minimum.
18. The method of claim 11 , wherein the pilot feedback indicators comprise instruments relevant to the current context, and
wherein the providing the one or more pilot feedback indicators comprises illuminating lights associated with the instruments relevant to the current context.
19. The method of claim 11 , wherein the pilot feedback indicators comprise proposed control actuations, and
wherein the providing the one or more pilot feedback indicators comprises displaying a proposed direction for a cockpit control to be actuated.
20. The method of claim 11 , wherein the pilot feedback indicators comprise a semantic label, and
wherein the providing the one or more pilot feedback indicators comprises displaying the semantic label on a display device in the cockpit.