SimEX: agents bootstrap robot skills in simulation, then fold towels in the real world
chris_j_paxton · x · 2026-11-10
RoboPapers Episode 116 covers SimEX by Jiaheng Hu, Steve Tod, and team (UT Austin), bringing LLM coding agents into real-world robotics.
- Problem: LLM agents increasingly handle complex long-horizon goals, but mostly on computer tasks; real-world robotics is far harder
- Approach: SimEX is a simulation-based auto-research framework — agents develop skills in simulation, building a robust toolbox transferable to real robots
- Results: acquired skills enable real tasks like towel folding and plate manipulation
A representative work on agents autonomously doing robotics research.
More from Embodied
- NVIDIA Ships Alpamayo 2 Edge: 2B-Parameter Onboard AV Reasoning at 160-190ms — nvidia · 2026-11-11
- NVIDIA launches open Alpamayo 2 Edge for real-time in-vehicle AI reasoning — nvidia · 2026-11-11
- NVIDIA open-sources GR00T 2 robot foundation model with 100K hours of training data — DrJimFan · 2026-11-11
- Stanford DroneHacks: 100 students fly GPS-denied PX4 drones on stereo cameras in 48 hours — mrpollo · 2026-11-10
- UT Austin open-sources OopsieVerse, a damage-aware safety benchmark for household robots — DynamicWebPaige · 2026-11-10
- Memories.ai pays users to record real-world tasks for robot training data — testingcatalog · 2026-11-10