Agent-Exploitation Affordances: From Basic to Complex Representation Patterns

Jul 16, 2026

Upload by Miles

In social robotics, agents must represent their action possibilities—affordances—to act effectively in real environments. While functional affordances have been widely studied through tool- and object-related knowledge representation, the impact of social context and other agents remains underexplored. This work introduces cooperative affordances, where interacting agents extend each other’s action ranges. It proposes a tractable ontological representation usable by artificial agents, and demonstrates the approach by composing elementary patterns to model diverse scenarios.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages13

Reading Time33 min

Introduction

Affordances and representation views

Capabilities and dispositions

Reciprocal and functional affordances

Social affordances and multi-agent action possibilities

arXiv :2607 .07475v 1 [ cs .RO] 8 Jul 2026 Agent-Exploitation Affordances: From Basic to Complex Representation Patterns Bastien Dussard 1[0009−0001−8800−8853], Aurélie Clodic 1[0009−0009−6484−8143], and Guillaume Sarthou 1[0000−0002−4438−2763] LAAS-CNRS, Université de Toulouse, CNRS, Toulouse, France [firstname.surname@laas.fr](firstname.surname@laas.fr) Abstract. In robotics, the capability of an artificial agent to represent the range of its action possibilities, i.e. affordances, is crucial to understand how it can act on its environment. While functional affordances, which refer to the use of tools and objects, have been broadly studied in knowledge representation, the implications of a social context and the presence of other agents have remained unexplored in this field. Consequently, in the field of social robotics, a multi-agent context enables the agents to engage in new actions that are potentially complementary to their individual capabilities, leading to the perspective of agentexploitation. This work focuses on the concept of cooperative affordance within the realm of social affordances. Cooperative affordances refer to situations where agents interact with each other to extend their action possibilities range. From this definition, this paper proposes a tractable ontological representation of this concept with the aim of making it usable by an artificial agent. Expanding on those elementary patterns, we illustrate the effectiveness of these representations by combining them to depict a diverse range of scenarios. Keywords: social affordances · knowledge representation · human-robot collaboration. 1 Introduction For an artificial agent to act efficiently upon its environment, it is essential that it possesses knowledge regarding its interaction capabilities with non-agentive entities, which enables it to bring about changes in its surroundings. Indeed, since for a robot to successfully complete its task requires that it understandshow to bring the environment to a desired different state, knowledge over the action possibilities it affords in a given environment is key. In the literature, this concept is commonly referred to as affordance and was first introduced by Gibson in [9] . Over the years, this concept has been refined in multiple fields given its open-to-interpretation nature, such as ecological psychology [20], industrial design [15], or robotics [17] . As each of those formalisms leveraged the concept for different purposes, they introduced refinements given their usage such as “disposition”,“perceivability”, or “effect” which 2 B. Dussard et al. captured important aspects. Nevertheless, two major views co-exist for representing affordances, either as a concept (Turvey in [23]) or as a relation between concepts (Chemero in [5]) . Those views differ given the author’s perspective, either environmental or agent/observer [17] . In this work, we build upon the formalism introduced by Chemero in [5], which outlines the relational nature of affordances. However, one could see that given the purpose of our work, it contradicts with the anti-representationalist perspective of most work in ecological psychology [6] . Nevertheless, as discussed in [26], “roboticists generally extract features as a basis for affordance detection and learning, thereby implicitly building an internal representation.”, motivating that the computational view over affordances is meaningful in robotics. Both Chemero’s original formalism and its extension for computational models consider affordances as emerging relations between agent’s capabilities and entities’dispositions. While a capability refers to “the ability to carry out a type of activity” [19], a disposition can be referred to as “the property of a thing that is a potential” [23] . Those two concepts work in pairs which enable action possibility. For instance, an agent capable of grasping and an object holding a graspable disposition can be matched t

FAQs

What problem does the paper address in social robotics?

It addresses how to represent an agent’s action possibilities when other agents and a social context influence what actions become available.

How does the paper define cooperative affordances?

Cooperative affordances are situations where agents interact with each other to extend the range of their possible actions.

What is the main contribution regarding representation?

The paper proposes a tractable ontological representation of cooperative affordances, designed to be usable by an artificial agent, and shows how to combine elementary patterns to depict varied scenarios.

Jul 16, 2026

Upload by Miles