Look into incorporating wheel motor current measurements for detecting and counteracting wheel slippage

[mkhlb]

Description of the task

Wheel motor current is directly proportional to torque. We can estimate how much torque we should be outputting (with perfect traction) and compare that with measured torque to detect when wheels are slipping.

I can think of 2 ways to do this.

(A): Which is employed by KIKS: using detection of motor slippage, we can alter the amount of power we give to motors to counteract slippage. This hinges on the assumption that if wheels are slipping in the previous timestep, they will slip in the next. (B): Which I have not seen yet in any TDPs: we can use the measured motor slippage to update kalman variances for a kalman filter running on predicting robot position/orientation from wheel velocities/IMU values, leading to more accurate on-robot position/orientation estimation inbetween precious vision packets.

See 4.1 for an example of someone doing (A)

See 3.1 for an KIKS doing (A)

Acceptance criteria

• [ ]

Blocked By