Towards Online Robot Interaction Adaptation to Human Upper-limb Mobility Impairments in Return-to-Work Scenarios

TL;DR

The paper tackles inclusive HRI for workers with upper-limb mobility impairments by presenting an online framework that integrates a subject-specific mobility model into a real-time augmented hierarchical quadratic program (AHQP).By modeling the human arm as a coupled, redundant chain and leveraging a Stack of Tasks (SoT), the approach optimizes a final interaction pose and the approach trajectory while enforcing joint- and task-space constraints and promoting residual mobility.Key contributions include the subject-specific impairment modelling with a diagonal impairment matrix, the Augmented HQP formulation that jointly plans human and robot motions, and a mobility-aware interaction strategy validated through simulations and multi-subject handover experiments.The work demonstrates improved comfort, reduced compensatory movements, and lower perceived workload compared with state-of-the-art ergonomic baselines, indicating potential for more inclusive and autonomous return-to-work scenarios.

Abstract

Work environments are often inadequate and lack inclusivity for individuals with upper-body disabilities. This paper presents a novel online framework for adaptive human-robot interaction (HRI) that accommodates users' arm mobility impairments, ultimately aiming to promote active work participation. Unlike traditional human-robot collaboration approaches that assume able-bodied users, our method integrates a mobility model for specific joint limitations into a hierarchical optimal controller. This allows the robot to generate reactive, mobility-aware behaviour online and guides the user's impaired limb to exploit residual functional mobility. The framework was tested in handover tasks involving different upper-limb mobility impairments (i.e., emulated elbow and shoulder arthritis, and wrist blockage), under both standing and seated configurations with task constraints using a mobile manipulator, and complemented by quantitative and qualitative comparisons with state-of-the-art ergonomic HRI approaches. Preliminary results indicated that the framework can personalise the interaction to fit within the user's impaired range of motion and encourage joint usage based on the severity of their functional limitations.

Figures (7)

• Figure 1: This example shows that optimising handover configurations only based on general ergonomic metrics (e.g. REBA hignett2000rapid) can force users with elbow impairments into awkward postures (shaded blue) or require whole-body compensation, like taking a step. By contrast, including mobility limitations in robot optimisation can improve user comfort (natural coloured posture) and overall interaction.
• Figure 2: Overview of the framework to model mobility limitations in impaired users and adapt the HRI strategy based on the severity of specific joints' impairment.
• Figure 3: Online adaptation results for various complete single-joint impairments, showing Rviz visualisation of initial and final configurations. From left to right, the plots depict optimised human joint angles, promoting residual mobility while respecting impaired RoM; robot ee desired (dashed) vs. actual (solid) positions and orientations, highlighting accurate tracking and compliance with task constraints; and the relative error between human and robot end-effectors, confirming successful establishment of interaction with different objects.
• Figure 4: Human redundancy resolution advised by AHQP optimisation for two conditions of partial multi-joint mobility impairments: elbow-dominant (left) and shoulder-dominant (right).
• Figure 5: Standing scenario with emulated elbow flexion impairment (EA): comparison of the REBA-defined transfer-point baseline (${\textit{SoA}}_\text{REBA}$) and the proposed mobility-aware handover.