cs.ROFeb 28, 2026

Test-Driven Agentic Framework for Reliable Robot Controller

Authors: Shivanshu Tripathi, Reza Akbarian Bafghi, Maziar Raissi

Organizations: Department of Electrical and Computer Engineering, University of California, Riverside · Department of Computer Science, University of Colorado, Boulder · Department of Mathematics, University of California, Riverside

Read PDF
Ask AI
Cite

Copy citation

DOINot available for this paper

Abstract

In this work, we present a test-driven, agentic framework for synthesizing a deployable low-level robot controller for navigation tasks. Given a 2D map with an image of an ultrasonic sensor-based robot, or a 3D robotic simulation environment, our framework iteratively refines the generated controller code using diagnostic feedback from structured test suites to achieve task success. We propose a dual-tier repair strategy to refine the generated code that alternates between prompt-level refinement and direct code editing. We evaluate the approach across 2D navigation tasks and 3D navigation in the Webots simulator. Experimental results show that test-driven synthesis substantially improves controller reliability and robustness over one-shot controller generation, especially when the initial prompt is underspecified. The source code and demonstration videos are available at: https://shivanshutripath.github.io/robotic_controller.github.io.

Figures & tables

Fig. 1: The figure shows map preprocessing pipeline. A user-provided map image is converted into a grayscale occupancy grid occ[y,x]occ[y,x] , where occ[y,x]=Trueocc[y,x]=\texttt{True} denotes an obstacle pixel and occ[y,x]=Falseocc[y,x]=\texttt{False} denotes free space. This pipeline simultaneously generates params.json , which stores the configuration parameters to produce occupancy grid.
Fig. 2: The figure shows the overview of our test-driven agentic framework. The workflow is partitioned in to three functional domains: (1) The red block shows the inputs to code_gen.py , providing preprocessed occupancy maps with params.json and the robot image; (2) The blue block highlights the calibration loop, where code_gen.py generates controller.py from prompt.py . PyTest suite evaluates the generated controller, and repair.py updates the prompt template or updates the code based on the failures; (3) The green block defines the PyTest files, which are run via repair.py .
Fig. 3: The figure shows the inputs to code_gen.py in the 33 D Webots simulation. The generator (LLM 1 ) produces controller.py conditioned on (i) the current prompt specification prompt.py (fixed base prompt plus accumulated updates), (ii) the Webots world file empty.wbt describing the arena, robot, goal, and static obstacles, and (iii) a human task prompt specifying start and goal poses. The optimizer (LLM 2 ) uses Pytest feedback to update prompt.py across iterations.
Fig. 4: The figure shows the performance of our approach in 22 D navigation described in Section IV-A . The left figure shows the success rate across learner models. The right figure shows cumulative success vs iteration kk .
Fig. 5: The figure shows the controllers performance for a 33 D navigation task in a Webots environment described in Section IV-B . The left figure shows the success rate across learner models. The right figure shows cumulative success vs iteration kk .
CategoryTest TargetVerification Objective
Static ContractComplianceEnsures required constants ( SENSOR_RANGE , etc.), imports, and main entry points are present.
 HygieneExplicitly bans forbidden libraries (e.g., scipy ) and unstable code patterns (e.g., indexing pygame surfaces).
Unit APIOccMap2DValidates boundary handling, out-of-bounds indexing, and occupancy querying logic.
 UtilityConfirms nearest_free correctly identifies navigable points for robot initialization.
 PlannerVerifies path format (Python int ), connectivity, and safety (no points in obstacles).
End-to-EndSchemaValidates return signatures of simulation loops for data logging and metrics calculation.
TABLE I: PyTests for 22 D map-based simulation
CategoryTest TargetVerification Objective
​​Static ContractComplianceEnsures required constants ( SENSOR_RANGE , etc.), imports, and main entry points are present.
 HygieneExplicitly bans forbidden libraries (e.g., scipy ) and unstable code patterns (e.g., indexing pygame surfaces).
Component LogicEnvironmentValidates environment parsing from world files.
 PlannerVerifies A∗A^{*} path connectivity, and ensures all planned waypoints lie strictly within free-space coordinates.
 ControlCheck accuracy of differential-drive wheel speed commands for forward motion, turning, and waypoint tracking.
End-to-EndSchemaValidates simulation by checking return schemas for positions, speeds, and collision counts.
TABLE II: PyTests for Webots simulation
Appendix figures & tables1 asset

Supplementary material from the paper’s appendix.

Appendix

Fig. 6: The figure shows the prompt-repair iterations in the 22 D simulation for a single run using GPT-5.2. Each iteration updates the prompt with new auto-repair rules derived from the previous failure, leading to a deployment-ready controller.
More in this topic

Robotic Control

Robotic control studies how machines use sensing and computation to produce reliable movement and interaction in changing environments.

See all 584 papers

Explore similar work

CardsList

  1. RAPID: Robot Agentic Programming from Demonstrations

    Sep 24, 2026Yuyao Liu, Jiayuan Mao, David Hsu +2Program SynthesisContact-Rich Robotic Manipulation

    Massachusetts Institute of Technology · National University of Singapore · University of Pennsylvania +1

  2. ModuLoop : Low-Level Code Generation using Modular Synthesizer and Closed-Loop Debugger for Robotic Control

    Jun 2, 2026Gina Yoon, Sumin Lee, Joo Yong SimRobotic ControlCode Generation for Robotics

    Department of Mechanical Systems Engineering, Sookmyung Women’s University, Cheongpa-ro 47-gil 100, Yongsan-gu, Seoul, 04310, South Korea

  3. DreamToNav: Generalizable Navigation for Robots via Generative Video Planning

    Mar 6, 2026Valerii Serpiva, Jeffrin Sam, Chidera Simon +4Robot NavigationVisual Robot Navigation

    The authors are with the Intelligent Space Robotics Laboratory, Skolkovo Institute of Science and Technology Moscow, Bolshoy Boulevard 30, bld. 1, 121205, Moscow, Russia.