[TRLC-DK1] Clock sync between robot PC and logger for builders labs (intermediate)
TRLC-DK1 discussion on clock sync between robot PC and logger, time drift, timestamp mismatches, and practical checks for reliable replay and datasets.
A surprisingly destructive DK1 data problem is simple time disagreement between machines: the robot PC and the logger both work, but their clocks drift enough to make replay and sensor alignment unreliable.
How are you validating clock synchronization between the robot PC and logging machines in DK1 workflows?
Please share how you detect drift, what synchronization method you trust, and how you confirm timestamps stay aligned during long sessions instead of only at startup.
If you reply, include one exact clock-drift symptom and one exact check or infrastructure change that improved time alignment.
The most reusable replies here explain how drift was detected over time rather than only how clocks were set once.
If you use a periodic sync validation during long capture sessions, share it. Searchers often need that operational guardrail.
It is also useful to say whether the fix was software config, network setup, or a changed logging architecture.