Giving Sense to Inputs: Toward an Accessible Control Framework for Shared Autonomy

TL;DR

The paper tackles the challenge of mapping 2D input to 6D robot motion in shared autonomy for assistive robotics by introducing a dynamic input mapping framework built on canal surface trajectories from GeoSACS. It demonstrates how aligning joystick corrections with canal disk axes, along with canal smoothing and controlled outside-canal motion, reduces user workload and improves usability compared to a GeoSACS baseline, validated in a 20-participant comparative study and an exploratory wheelchair-user study. Key contributions include the final mapping framework with explicit geometric constructs, insights from pilot work, and practical design considerations for accessibility, paving the way for real-world deployment and participatory design with disabled users. The work has significant implications for making high-DoF assistive robots more usable and controllable by people with mobility limitations, and it provides a concrete methodology and empirical evidence to guide future development and deployment.

Abstract

While shared autonomy offers significant potential for assistive robotics, key questions remain about how to effectively map 2D control inputs to 6D robot motions. An intuitive framework should allow users to input commands effortlessly, with the robot responding as expected, without users needing to anticipate the impact of their inputs. In this article, we propose a dynamic input mapping framework that links joystick movements to motions on control frames defined along a trajectory encoded with canal surfaces. We evaluate our method in a user study with 20 participants, demonstrating that our input mapping framework reduces the workload and improves usability compared to a baseline mapping with similar motion encoding. To prepare for deployment in assistive scenarios, we built on the development from the accessible gaming community to select an accessible control interface. We then tested the system in an exploratory study, where three wheelchair users controlled the robot for both daily living activities and a creative painting task, demonstrating its feasibility for users closer to our target population.

Figures (4)

• Figure 1: In this shared autonomy framework, after two demonstrations, the motion is encoded in a canal surface. While the robot autonomously navigates within the canal, users can provide corrections with a gamepad, aligned to the correction axes on the canal’s cross sections to ensure intuitive mapping.
• Figure 2: Categories of disks found within a canal
• Figure 3: Data collection from participants including the generated canals for the three tasks.
• Figure 4: Subjective results from the comparative study conducted with 20 participants.