[DK1] Bimanual bringup and camera sync for builders labs (advanced)
Advanced DK1 discussion around bimanual bringup, camera synchronization, and the failure patterns that appear when experiments scale.
Forum hub for TRLC-DK1 setup, calibration, LeRobot logging, bimanual bringup, and camera sync discussion. 22 posts.
Advanced DK1 discussion around bimanual bringup, camera synchronization, and the failure patterns that appear when experiments scale.
DK1 getting started discussion covering installation, first bringup, and calibration habits for builders and labs.
Discussion thread on DK1 LeRobot data collection workflow, logging discipline, and experiment structure for builders and labs.
TRLC-DK1 discussion on action and video counts not matching after batch export, silent dataset corruption risks, and how labs audit exports before training.
TRLC-DK1 discussion on validators passing while one camera stream is still shorter, hidden temporal inconsistencies, and advanced audit strategies.
TRLC-DK1 discussion on camera device order changing after reboot, capture scripts binding to the wrong sensors, and how labs keep camera assignments stable.
TRLC-DK1 discussion on camera timestamp drift, sync validation, capture alignment, and practical checks for usable robot-learning datasets.
TRLC-DK1 discussion on clock sync between robot PC and logger, time drift, timestamp mismatches, and practical checks for reliable replay and datasets.
TRLC-DK1 discussion on dual-arm frame alignment, handover accuracy, coordinate consistency, and calibration checks for bimanual setups.
TRLC-DK1 discussion on episode boundaries, bad-demonstration filtering, dataset QA, and practical checks before uploading or training.
TRLC-DK1 discussion on delayed stop events, truncated rollouts, partial exports, and how labs confirm clean episode endings before using data.
TRLC-DK1 discussion on exports that contain videos but miss metadata JSON, partial export failure, and audit checks before dataset use.
TRLC-DK1 discussion on export scripts succeeding while one episode folder is empty, silent partial failures, and checks to catch them before use.
TRLC-DK1 discussion on failure replay, bad episode triage, debugging strategy, and how labs decide what to keep, relabel, or discard.
TRLC-DK1 discussion on handoff notes claiming capture is good while the next shift's validator fails, and better transition checks for lab teams.
TRLC-DK1 and LeRobot discussion on camera calibration, reprojection error, multi-view consistency, and when calibration is good enough to trust.
TRLC-DK1 discussion on logging starting before cameras are ready, empty opening seconds in episodes, and how labs enforce startup readiness gates.
TRLC-DK1 discussion on metadata mismatch between actions, video, and annotations, plus how labs audit dataset consistency before training.
TRLC-DK1 discussion on packet loss during remote teleoperation and logging, unstable network paths, and how labs protect data quality under network stress.
TRLC-DK1 discussion on SSD write speed, logging bottlenecks, session loss during data capture, and how labs validate storage reliability before runs.
TRLC-DK1 discussion on USB camera bandwidth, dropped frames, bus saturation, and practical multi-camera capture fixes for labs.
TRLC-DK1 discussion on USB hub power cycles silently changing camera FPS, capture quality drifting, and how labs verify stream settings after resets.