0:00:34 New Year's Update and Marker Cap Success: The stream marks the beginning of the 2025 season. The new black marker caps, designed to protect motion-tracking markers from the cats, are surprisingly effective.
0:02:58 Cable Issue: A loose cable caused one of the robot's legs to malfunction, highlighting a major flaw in the error handling system. The cable had become uncrimped and slipped out of its connector.
0:05:06 Error Handling Overhaul: The error handling system was revamped to store multiple errors instead of just one. This allows for a more robust response to issues. However, the specific error caused by the loose cable may still not be detectable without manual inspection.
0:08:29 Catching Progress: JuggleBot can now "catch" a ball, although it's more of a technicality due to the ball bouncing off a tube before landing in the hand. This demonstrates successful integration of the tracking system and robot movement.
0:10:47 Coordinate System Discrepancy: A 90-degree difference between the tracking software's coordinate system and the robot's internal coordinate system caused initial movement errors, but was resolved by rotating the tracking data.
0:14:03 Leg Collapse Issue: During catch attempts, one leg consistently collapses upon returning to the flat position. The cause is uncertain but might be related to torque limits or the platform's aggressive tilt.
0:18:15 Tracking Marker Bisected: When the ball is in the hand, the tracking software sometimes registers it as two balls due to the hand's carbon fiber tubes bisecting the markers. A potential solution is to ignore tracking data when the ball is expected to be in the hand.
0:23:39 Boot Sequence Simplification: The boot sequence was simplified by automatically polling the last known state (homing, encoder search, moving) and bundling this information into the robot state topic, eliminating unnecessary complexity.
0:26:30 Data Logging with PlotJuggler: PlotJuggler is now being used to record and analyze robot data, providing valuable insights into the robot's behavior during operation. This will help diagnose issues and improve performance.
0:34:53 Hand Control Update: The hand control is being updated to use a single message for commanding position, velocity, and torque, simplifying the control process.
0:37:53 O-Drive Gain Scaling Issue: Difficulties were encountered when changing the O-Drive's gain scaling, with the hand behaving unexpectedly despite setting seemingly correct values. This requires further investigation.
0:50:44 Testing Rig Challenges: The testing rig designed to safely test the hand's functionality has a limitation: the O-Drive can only pull the hand up, not down, leading to cable tangling issues.
0:56:23 Motor Spool Unscrewing: Due to the hand occasionally hitting the top stop, the motor spool on JuggleBot is unscrewing itself, requiring manual tightening and potentially leading to a more permanent solution like a metal part or a lock nut.
0:59:33 Catching Strategy: The presenter argues that a passive catching strategy, where the hand is compliant and moves with the ball, might be more effective than trying to perfectly match the ball's velocity.
1:05:19 Spring-Damper System: A spring-damper system is being explored to make the hand compliant during catching. The goal is to have the hand smoothly absorb the ball's impact and stop at a defined position.
1:09:33 O-Drive Gain Calculation: The current challenge is to translate the theoretical spring-damper system parameters into appropriate O-Drive position and velocity gains.
1:16:21 Hand Behavior Inconsistencies: The hand's behavior is inconsistent, sometimes moving as expected and sometimes not, even with the same settings, making it difficult to determine the correct gain values.
1:34:53 Future Plans: The presenter plans to continue investigating the O-Drive gain issue, improve data logging, and further develop the passive catching strategy.