Robotics
Research on Robotics: explanations, evidence, limitations and connections between papers.
16 results
A mapless depth-image query speeds reactive drone collision checks
A GPU method queries depth images directly, reducing collision-check latency in a tested 400-query drone-planning setting.
One robot model learns to predict scenes and control varied robots
GreenWAM’s robot-specialized starting point improves RoboTwin task success over its parent under a fixed adaptation recipe.
Trace a robot’s prediction failure to the missing evidence
A single camera frame can miss motion. After deployment, changed conditions can make a learned account of action consequences unreliable. A model can also predict plausible outcomes while failing to distinguish actions that succeed from actions that fail. The three studies point to different evidence for these problems: recent visual history, valid deployment transitions, and executed action alternatives with verified outcomes. Their tests measure different behaviors, so each result supports a bounded choice of evidence.
A driving planner organizes candidate paths by motion
ManeuverFlow groups driving paths by motion and uses shared future-scene evidence to score and refine them.
Robot models improve when actions explain intended effects
Arc adds causal traces to existing robot demonstrations and improves control across tested simulation and hardware tasks.
World-model rehearsal helps robots retain skills while learning
D2R helps robots learn sequential tasks while retaining earlier skills through regenerated world-model rehearsal.
Reflex helps a humanoid catch thrown boxes with moving-camera vision
Reflex separates body control, box-motion prediction, and vision so a humanoid can catch thrown boxes, with long-range limits.
Why Robots Forget Skills, Choose Weak Paths, or Stumble Over Words
A robot may lose an earlier skill after learning a new one, choose poorly among possible paths, or fail when a familiar request is worded differently. Each symptom points to a different place where useful evidence could enter its decisions. The studies test rehearsal during learning, predicted scene evidence during path selection, and a one-time instruction rewrite before action. Each reports an outcome within its own task and evaluation; the value of the intervention depends on the failure it addresses.
GLIO2 completes a route with weak LiDAR geometry
GLIO2 jointly updates LiDAR, inertial, and satellite measurements, completing a tested degenerate bridge route where LiDAR systems diverged.
Human motion helps humanoid robots follow instructions
VioLA uses shared human and robot motion targets to train humanoid control, with strong results in one laboratory evaluation.
Rewarding diverse robot successes improved tested shift averages
DRIVE rewards diverse successful robot behavior; simulation and two physical tasks show higher average success under tested shifts.
Juno connects robot predictions to motion and action learning
Juno links motion-aware robot predictions to action learning, with gains in tested simulation tasks and physical shifts.
Robot history helps most when video learning predicts forward
Robot tests find longer visual history helps after causal video pretraining, while bidirectional pretraining shows no net gain.
RealtimeWAM speeds robot action prediction while keeping success
RealtimeWAM coordinates computation in robot action models, lowering average simulation latency while keeping task success close to original inference.
Rules help frozen robot models handle new wording
A rule-based rewriter improved a frozen robot model on new instruction phrasings in simulated tasks.
Robot models learn from executed action alternatives to predict failures
CureWM repairs released robot world models with executed action alternatives, improving success–failure prediction in several tested settings.