[Paper Review] Human-Robot Collaboration: From Psychology to Social Robotics

Judith Bütepage, Danica Kragić|arXiv (Cornell University)|May 29, 2017

Social Robot Interaction and HRI15 citations

TL;DR

This paper proposes an embodied approach to human-robot collaboration (HRC) inspired by human-human interaction (HHI) psychology, emphasizing sensorimotor contingencies for mutual, adaptive collaboration. By modeling shared representations and dynamic coordination through sync SMCs (sensorimotor contingencies), the framework enables autonomous, physically interactive HRC without leader-follower assumptions, advancing trust and efficiency in joint tasks.

ABSTRACT

With the advances in robotic technology, research in human-robot collaboration (HRC) has gained in importance. For robots to interact with humans autonomously they need active decision making that takes human partners into account. However, state-of-the-art research in HRC does often assume a leader-follower division, in which one agent leads the interaction. We believe that this is caused by the lack of a reliable representation of the human and the environment to allow autonomous decision making. This problem can be overcome by an embodied approach to HRC which is inspired by psychological studies of human-human interaction (HHI). In this survey, we review neuroscientific and psychological findings of the sensorimotor patterns that govern HHI and view them in a robotics context. Additionally, we study the advances made by the robotic community into the direction of embodied HRC. We focus on the mechanisms that are required for active, physical human-robot collaboration. Finally, we discuss the similarities and differences in the two fields of study which pinpoint directions of future research.

Motivation & Objective

• To address the lack of reliable human and environmental state representation in HRC, which limits autonomous decision-making.
• To overcome the leader-follower bias in current HRC systems by modeling collaboration as mutual, interdependent action.
• To integrate psychological findings on human-human interaction (HHI) into robotic systems for more natural, adaptive collaboration.
• To develop a framework for active, physical HRC based on embodied intelligence and sensorimotor coupling.
• To identify design principles for HRC systems that support shared goals, mutual adaptation, and reduced cognitive load.

Proposed method

• Adapts psychological findings on sensorimotor contingencies (SMCs) from HHI to HRC, distinguishing check SMCs (unidirectional) and sync SMCs (mutual coupling).
• Models sync SMCs as dynamic, reciprocal sensorimotor loops enabling real-time coordination and joint action in physical tasks.
• Applies embodied intelligence principles to HRC, where robots perceive and act based on continuous, context-sensitive interaction with humans.
• Uses shared representations of the environment and task goals to enable interdependent action sequences in collaborative scenarios.
• Integrates anticipatory and adaptive behaviors through intention modeling and multimodal sensing (e.g., motion, gaze, gesture).
• Evaluates robotic systems using benchmarks from HRI, including joint task execution, handover motions, and collaborative manipulation.

Experimental results

Research questions

• RQ1 How can sensorimotor contingencies (SMCs) from human-human interaction be modeled and applied to enable autonomous human-robot collaboration?
• RQ2 What mechanisms are required for robots to actively engage in joint actions without assuming a leader-follower role?
• RQ3 How can shared representations of the environment and task goals be maintained and updated in real time during physical collaboration?
• RQ4 What role do embodied, perceptually grounded behaviors play in reducing human cognitive load and improving collaboration efficiency?
• RQ5 How can robots anticipate human actions and adapt their behavior in real time to support seamless collaboration?

Key findings

• Sync SMCs—mutual sensorimotor coupling—enable more effective and fluid collaboration than unidirectional or leader-follower models.
• Active, embodied collaboration reduces human cognitive workload and increases task efficiency compared to reactive or instruction-based HRI.
• Robots that model human intentions and adapt in real time through sensorimotor feedback show improved fluency and perceived teamwork quality.
• Physical collaboration is most effective when robots share representations of the environment and task goals, enabling mutual adaptation and trust.
• The integration of psychological principles of HHI into robotics leads to more natural, intuitive, and autonomous HRC systems.
• Evaluation in joint manipulation and handover tasks confirms that sync SMC-based systems outperform traditional reactive or pre-programmed approaches in collaboration quality and robustness.