← Back to Model Beat
Research·Oct 12·all news from October 12, 2026

Learning to Play Like a Pro: Adaptive Motion Planning and Tracking for Humanoid Robots

Roboticists have released a white paper detailing a method to decouple motion planning from motion tracking, allowing humanoid robots to better replicate professional athletic movements. By employing adaptive, speed-aware control, this approach aims to reduce the performance discrepancy between simulated robotic training and real-world execution.

Covered by 1 source

Related stories

ResearchOpenAI's safety crisis keeps getting worse and the company keeps making it worseOct 8 · 10 sourcesResearchGoogle says self-improving AI isn't anywhere closeOct 14 · 2 sourcesResearchAnthropic gives METR inside access to audit its AI models and safety practicesOct 15 · 2 sourcesResearchPlan or Retrieve First? A Scale Study of Tool Discovery ArchitecturesOct 15 · 3 sources