Why does this policy fail outside simulation?
98.4% in sim, 51% on hardware, same seeds. It turned out to be actuator latency: our sim assumed 2 ms, the real arm is 18 ms under load. Posting the identification procedure.
#sysid#latency
98.4% in sim, 51% on hardware, same seeds. It turned out to be actuator latency: our sim assumed 2 ms, the real arm is 18 ms under load. Posting the identification procedure.
Latency is the first thing to check and the last thing anyone checks. Log commanded vs measured joint position and look at the phase lag.
We found the same, plus a 6 ms jitter from DDS discovery. Pinning the participant removed it.