The Law of Task-Achieving Body Motion: Axiomatizing Success of Robot Manipulation Actions

TL;DR

The paper addresses the lack of guarantees in open-world robot manipulation by proposing the Law of Task-Achieving Body Motion, an axiomatic correctness framework that ties task requests to physical execution within scoped physics. It introduces Task--Environment--Embodiment (TEE) classes and Semantic Digital Twins (SDTs) to formalize state, dynamics, and validity, and defines three verifiable predicates—$SatisfiesRequest$, $Causes$, and $CanPerform$—to certify task achievement via a single CanAchieve condition. The framework enables motion synthesis, verification, and diagnostic reasoning (including counterfactuals) while clearly delimiting scope through physics validity intervals $I_{\Phi}$. The authors instantiate the approach on articulated container manipulation in kitchen settings across three mobile manipulators, demonstrating practical use modes: generation, verification, learning from observations, and cross-embodiment analysis. The work advances a modular, certifiable approach to combining semantics, causality, and embodiment in robotic manipulation, with future directions toward deformable objects, more realistic physics, and integration with foundation models for generated task plans and verifications.

Abstract

Autonomous agents that perform everyday manipulation actions need to ensure that their body motions are semantically correct with respect to a task request, causally effective within their environment, and feasible for their embodiment. In order to enable robots to verify these properties, we introduce the Law of Task-Achieving Body Motion as an axiomatic correctness specification for body motions. To that end we introduce scoped Task-Environment-Embodiment (TEE) classes that represent world states as Semantic Digital Twins (SDTs) and define applicable physics models to decompose task achievement into three predicates: SatisfiesRequest for semantic request satisfaction over SDT state evolution; Causes for causal sufficiency under the scoped physics model; and CanPerform for safety and feasibility verification at the embodiment level. This decomposition yields a reusable, implementation-independent interface that supports motion synthesis and the verification of given body motions. It also supports typed failure diagnosis (semantic, causal, embodiment and out-of-scope), feasibility across robots and environments, and counterfactual reasoning about robot body motions. We demonstrate the usability of the law in practice by instantiating it for articulated container manipulation in kitchen environments on three contrasting mobile manipulation platforms

Figures (3)

• Figure 1: The Law of Task Achieving Body Motion(middle box) axiomatizes the success of robot manipulation actions under a scoped physics model (top left box) and by splitting success validation into the dimensions of semantic correctness, causal sufficiency and embodiment feasibility (top right box).
• Figure 2: Motions $\tau$ (red lines) generated by evaluating $\mathit{SatisfiesRequest}(\text{open()}, G_{\text{final}}) \wedge \mathit{Causes}(\tau, G_{\text{final}}, \Phi, I_{\Phi})$ for the kitchen B (top) and kitchen A (bottom) in distinct states. The left sides shows the results when all containers are initially closed. The right side shows the results when some containers are randomly selected to be open. No motion exists for containers that are already open to become open.
• Figure 3: The left side shows a CanPerform() failure of the Stretch robot for the task open(oven door). Due to its limited reachability the Stretch is not able to perform the opening motion that starts with the gripper at the handle. The right side shows the PR2 being able to perform that task.

Theorems & Definitions (3)

• Definition 1: Manipulation Task Specification
• Definition 2: Task--Environment--Embodiment Class
• Definition 3: Semantic Digital Twin