aiqbench analysis
This research introduces a novel evaluation metric for robot co-design that assesses morphology competence by restricting the controller to low-complexity families, revealing that standard task rewards often fail to capture intrinsic design quality. The study demonstrates that this restricted-control approach provides a more reliable and interpretable property of a robot's body than geometric descriptors or raw reward scores alone. By decoupling morphological potential from specific control policies, this method offers a critical new lens for evaluating embodied AI systems beyond simple performance benchmarks.
Beyond Task Reward: A Controller-Restriction Protocol for Evaluating Embodiment-Dependent Competence
arXiv:2610.07629v1 Announce Type: new Abstract: Co-design methods optimize a robot's body and controller jointly and judge the result by one number, the task reward of the fully optimized pair. That number cannot separate morphologies whose competence depends on the controller to very different degrees. We evaluate a morphology by restricting its controller instead, recording the task competence it retains under an explicitly declared, low-complexity controller family, environment, task and search budget. On three EvoGym locomotion tasks, task reward explains only $39\%$, $33\%$ and $10\%$ of the variance in this quantity, and geometric descriptors do not predict it under run-grouped cross-validation. The measurement is reliable across optimizer restarts (ICC$(2,k) = 0.956$--$0.986$) but depends on the declared family: phasing the drive by actuator index instead of position ranks the same morphologies at Spearman $0.50$--$0.63$ and reverses reward-matched pairs. As a second search objective the axis improved competence at matched task reward in $3$ of $5$ paired runs, short of a pre-registered bar of $4$. Used after an ordinary reward-only search instead, to choose within its top task-reward band, it selected a different body in all $8$ runs offering a choice, at a cost of at most $0.10$ reward units, and in $6$ of $8$ that body also scored higher under a held-out family. Restricted-control competence is therefore a reportable property of a co-designed morphology, interpretable only with the controller family that defines it.
Abstract:Co-design methods optimize a robot's body and controller jointly and judge the result by one number, the task reward of the fully optimized pair. That number cannot separate morphologies whose competence depends on the controller to very different degrees. We evaluate a morphology by restricting its controller instead, recording the task competence it retains under an explicitly declared, low-complexity controller family, environment, task and search budget. On three EvoGym locomotion tasks, task reward explains only $39\%$, $33\%$ and $10\%$ of the variance in this quantity, and geometric descriptors do not predict it under run-grouped cross-validation. The measurement is reliable across optimizer restarts (ICC$(2,k) = 0.956$--$0.986$) but depends on the declared family: phasing the drive by actuator index instead of position ranks the same morphologies at Spearman $0.50$--$0.63$ and reverses reward-matched pairs. As a second search objective the axis improved competence at matched task reward in $3$ of $5$ paired runs, short of a pre-registered bar of $4$. Used after an ordinary reward-only search instead, to choose within its top task-reward band, it selected a different body in all $8$ runs offering a choice, at a cost of at most $0.10$ reward units, and in $6$ of $8$ that body also scored higher under a held-out family. Restricted-control competence is therefore a reportable property of a co-designed morphology, interpretable only with the controller family that defines it.
Robotics (cs.RO)
arXiv:2610.07629 [cs.RO]
(or arXiv:2610.07629v1 [cs.RO] for this version)
https://doi.org/10.48550/arXiv.2610.07629 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 319 words
Related coverage
[AINews] Reality Checks on AI News (Yegge shuts down Gas Town, Databricks’ +60% Astra cost)
Latent Space · Sep 17, 2026
[AINews] Meta Muse Wonder & Wonder Full: MSL’s Claude 5.5-tier Frontier Model
Latent Space · Oct 13, 2026
Studying the role of Sandboxing for AI Control
LessWrong · Jul 17, 2026
MSK-Bench: Benchmarking Full-Body Musculoskeletal Motor Control Across Tasks, Control Paradigms, and Physiological Metrics
arXiv · Sep 24, 2026