Compliant force sensing system
Sensing systems as well as their methods of operation and training are described. In some embodiments, a sensing system may include a compliant contact pad configured to contact an environment, and a plurality of sensors configured to detect a physical parameter associated with deformation of the compliant contact pad. A processor configured to receive signals from the plurality of sensors may determine a magnitude and direction of a force applied to the compliant contact pad.
1. A sensing system comprising:
a compliant contact pad configured to contact an environment;
a plurality of sensors coupled to the compliant contact pad, wherein the plurality of sensors are configured to detect a physical parameter indicative of deformation of the compliant contact pad, and
a processor configured to receive signals from the plurality of sensors, wherein the processor is configured to determine a magnitude and direction of a force applied to the compliant contact pad with the signals from the plurality of sensors.
2. The sensing system of claim 1 , wherein the processor is configured to determine a contact location of the force applied to the compliant contact pad with the signals from the plurality of sensors.
3. The sensing system of claim 1 , wherein the plurality of sensors is at least three sensors.
4. The sensing system of claim 1 , wherein the plurality of sensors is at least five sensors.
5. The sensing system of claim 1 , wherein an exterior surface of the compliant contact pad is hemispherical or convex.
6. The sensing system of claim 1 , wherein the compliant contact pad is an elastomer.
7. The sensing system of claim 1 , wherein the plurality of sensors include at least one selected from the group of force sensors, pressure sensors, torque sensors, strain gauges, and displacement sensors.
8. The sensing system of claim 1 , further comprising a rigid support, wherein the compliant contact pad is disposed against the rigid support.
9. The sensing system of claim 8 , wherein the plurality of sensors are disposed between the compliant contact pad and the rigid support.
10. The sensing system of claim 8 , wherein the plurality of sensors are disposed in cavities formed in the rigid support.
11. A system comprising:
the sensing system of claim 1 ,
wherein the processor is configured to determine a contact location of the force against the compliant contact pad with the signals from the plurality of sensors, wherein the processor is configured to control at least one operation of the system based at least partly on the determined magnitude, direction, and/or contact location of the force.
12. The system of claim 11 , wherein the system comprises at least one selected from the group of a robotic limb and a robotic hand.
13. The system of claim 11 , wherein the sensing system is disposed on a distal portion of a robotic limb segment of the system.
14. A method comprising:
applying a force to a compliant contact pad;
receiving signals from a plurality of sensors configured to detect a physical parameter indicative of deformation of the compliant contact pad; and
providing the signals to a trained statistical model of the compliant contact pad and obtaining a corresponding output including a magnitude and direction of the force applied to the compliant contact pad.
15. The method of claim 14 , further comprising controlling at least one operation of a system based at least partly on the output magnitude and direction of the force.
16. The method of claim 14 , wherein the output includes a contact location of the force applied to the compliant contact pad.
17. The method of claim 14 , wherein the plurality of sensors are configured to detect a stress state of the compliant contact pad.
18. The method of claim 14 , wherein the plurality of sensors are configured to detect a strain state of the compliant contact pad.
19. A method comprising:
obtaining training data, wherein the training data include magnitude, direction, and contact location data for forces applied to a compliant contact pad, and wherein the training data includes sensor data from a plurality of sensors configured to detect a physical parameter indicative of deformation of the compliant contact pad when the forces are applied to the compliant contact pad;
generating a trained statistical model using the training data; and
storing the trained statistical model in a non-transitory computer readable medium for subsequent use.
20. The method of claim 19 , wherein generating the trained statistical model includes using regression.
21. The method of claim 20 , wherein using regression includes using Gaussian process regression.
22. The method of claim 19 , wherein generating the trained statistical model includes using a neural network.