Self-evolving runtime infrastructure for Physical AI and embodied agents. Ground AI agents into robot bodies with e-URDF, sandbox safety, capability routing, praxis capture, physical memory, runtime intervention, and skill evolution.

Implementation of SHARP: Shielding-Aware Robust Planning for Safe and Efficient Human-Robot Interaction - RAL 2022

Synthesis of control barrier functions with SOS

ABC-DS: obstacle Avoidance with Barrier-Certified polynomial Dynamical Systems

A next-gen ESP32-C3 daemon for real-time robot monitoring, safety control, and ROS2 integration.

Synthesizing safe robot policies in joint physical-belief spaces with deep RL! - CoRL 2023

Formal adversarial testing of LLM-generated industrial robot (URScript) code: ISO 10218-1:2025 safety + CWE security analysis, AST static watchdog, URSim runtime. Simulation-only.

Open-source evidence layer for robots. ROS 2 incident recording, signed manifests, MCAP export, and safe replay.

Real-time robot collision detection using supervised learning methods

Curated standards, runtime controls, assurance methods, and evaluation resources for governing embodied and physical AI.

ForeForce mmWave cobot safety system for IWR6843AOP and IWRL6432AOP

A ROS Noetic package for sensor-based collision prevention in mobile robots. Supports both 2D (STDR) and 3D (Gazebo) simulations with configurable safety layers to stop or adjust motion before impact.

ROS-based collision prediction system for TIAGo robot manipulation, using Python perception and trajectory analysis to anticipate contact risks and improve safe autonomous operation.

Robot incident evidence validation and reconstruction CLI

Asimov's Three Laws compiled to real Cedar authorization policy that acts as the runtime safety layer for AI agents which touch the physical world. Governs the real SO-101 (LeRobot) arm over MCP.

Deterministic certification platform — kernel + N domain packs (4th of the -stra family; the first platform emitted by the HELIX Condense recipe from the robotics/release cluster)

Policy Library CBF: A runtime safety filter based on parallel rollouts of multiple fallback policies with a computation overhead of < 10 ms.

Reference VLA action gate for validating robot commands before physical execution.

Safe hierarchical embodied-AI framework with multi-layer safety filtering and formal recovery — IEEE TENSYMP 2026