Robot agent harness ensures verified task completion with self improvement
RegenHarness: A Robot Agent Harness with Evidence-Gated Recursive Self-Improvement
Robotics
Summary
Robots often struggle to clearly know when they finish tasks correctly. The authors created a system called RegenHarness that helps robots plan, act, observe, and verify their work step-by-step, keeping track of what’s been done and what can be trusted. It even lets the system improve itself safely by making controlled, tested changes over time without risking mistakes. They tested this on a real four-legged robot doing tasks like warehouse navigation and reporting, showing it works in complex, real-world situations.
What this means in practice
- •For robotics development teams: Deploy robots that reliably verify task completion and autonomously improve their execution strategies in complex environments.
- •For warehouse operations teams: Use quadruped robots to perform voice-directed navigation, inspection, delivery, and reporting with reliable status verification.$Commercial implications: Enables commercial deployment of reliable, multi-skill robots for warehouse management with audit trails and task verification.
Authors
Kailin Wang, Haoxiang Jie, Yaoyuan Yan, Zhiyou Heng, Zhaosong Li
Abstract
Long-horizon robot execution requires a clear distinction between a model's proposal, a controller's termination, and verified task completion. We present RegenHarness, an evidence-gated robot-agent harness connecting task planning to heterogeneous robot skills. Its execution architecture couples a model loop for context-conditioned proposals with an agent loop for dispatch, observation, verification, commitment, and bounded recovery. Four role-isolated contexts separate planning, supervision, verification, and recovery inputs. Versioned memory distinguishes observed facts from accepted task progress, while an identity- and version-bound commit gate controls updates to trusted task state. The runtime combines duplicate-dispatch control, resource leases, and recovery budgets under explicit backend contracts, and checks the original user goal before reporting completion. To our knowledge, we are the first to introduce an evidence-gated recursive self-improvement (RSI) protocol for embodied robotic agents. Across missions, execution records motivate candidate changes to context rules, task templates, routing, and recovery policies; fixed regression checks and release authorization govern their acceptance; versioned rollout and rollback preserve configuration traceability. This RSI protocol revises the harness configuration without online model-weight updates or permission to weaken the commit gate. A real quadruped deployment documents voice-triggered warehouse navigation, panoramic inspection, visual analysis, message delivery, return, and spoken reporting through linked audio, images, trajectories, and receipts. A separate circuit demonstrates why completion depends on execution history rather than endpoint proximity alone. Together, the cases demonstrate integrated perception, physical execution, communication, and history-dependent completion in real-world robot tasks.