3 Franka arms stack 9-block tower with Gemini Robotics-ER as VLA brain
Zergylord · x · 2026-09-19
An open-source project, IsaacSimGeminiRobotics, demonstrates multi-robot cooperation: three Franka FR3 arms build a 9-block tower inside NVIDIA Isaac Sim with ROS 2, orchestrated by Google Gemini Robotics-ER acting as a Vision-Language-Action (VLA) brain.
- Architecture: Gemini Robotics-ER analyzes overhead camera feeds, plans tasks, and dispatches parallel robot commands via Function Calling
- Dual modes: switch between the Gemini VLM orchestrator and a rule-based motion planner that needs no API
- Engineering insight: the model produces "ghost grasps" because it can't see a missing attach frame; gating tools on TF plus a precondition check eliminated them — "plan → execute → write the limit back into the machine spec"
- Monitoring: a Salabim-inspired React dashboard shows live KPI utilization, a 2D digital twin map, timelines, and discrete event logs
- Cross-OS: Windows ↔ WSL2 bridged via FastDDS Unicast, automated at boot
Related event: Three Franka Robot Arms Stack Tower Collaboratively with Gemini Robotics-ER(2 posts)→
More from coding & agent
- Human Oversight Is 70-75% of AI Agent Costs in Banking, McKinsey Finds — mikeflache · 2026-11-04
- Over 80% of Major Financial Institutions Now Run Agentic AI in Production — mikeflache · 2026-11-04
- Mnemos MCP nears launch after shared memory layer fix — RileyRalmuto · 2026-11-04
- Moving mobile dev to remote machines makes it multiplayer across product, design, engineering — rudrank · 2026-11-04
- Kuaishou's EvoRec: self-evolving agentic recommender lifts ad revenue 4.6% — _reachsumit · 2026-11-04
- TRIO: three specialized agents with aligned objectives refine retrieval-augmented generation — _reachsumit · 2026-11-04