What problem does it solve?
Automates end-to-end robotics idea discovery workflow, starting from literature review and ending with concrete, benchmark-grounded ideas for embodied AI, reducing time to actionable concepts.
Core Features & Use Cases
- Phase 0: Frame the Robotics Problem from user input and project context.
- Phase 1: Robotics Literature Survey using /lit-survey focused on CoRL/RSS/ICRA/IROS/RA-L milestones and recent arXiv work, to build a robotics-centric landscape.
- Phase 2: Robotics-Specific Idea Generation and Filtering that passes the Robotics Problem Frame and landscape matrix into the idea generator, producing top candidates with clear framing and metrics.
- Phase 3: Feasibility and Pilot Design outlining a minimal validation plan (sim/offline) before any real hardware, including baselines and success/failure criteria.
- Phase 4: Deep Novelty Verification via /lit-novelty-check with explicit embodiment, task, benchmark, sensor stack, and sim2real considerations.
- Phase 5: External Robotics Review via /research-review to gather expert critique on novelty and practicality.
- Phase 6: Final Report preparation for downstream workflows.
Quick Start
Describe your robotics direction and run the idea discovery workflow.