Probabilistic vs linear blending approaches to shared control for wheelchair driving

Area of Science:

• Robotics
• Human-Computer Interaction
• Rehabilitation Engineering

Background:

• Individuals with severe mobility impairments face challenges operating conventional powered wheelchairs.
• Smart wheelchairs offer solutions through assistive technologies and alternative interfaces.
• Shared control strategies aim to balance user autonomy with system assistance.

Purpose of the Study:

• To implement and evaluate a generalized form of shared control for smart wheelchairs.
• To compare the effectiveness of probabilistic shared control (PSC) against linear blending (LB).
• To assess the impact of PSC on user safety, particularly for novice users.

Main Methods:

• Implementation of probabilistic shared control (PSC) for smart wheelchairs.
• Comparison of PSC with traditional linear blending (LB) techniques.
• Evaluation of safety and user interaction within a shared control framework.

Main Results:

• Probabilistic shared control (PSC) demonstrates improved accuracy in modeling user-wheelchair interaction.
• PSC shows practical success over linear blending (LB) in enhancing safety.
• Novice users particularly benefit from the enhanced safety provided by PSC.

Conclusions:

• Probabilistic shared control offers a more robust approach to smart wheelchair assistance.
• PSC's ability to model interaction uncertainty leads to superior safety outcomes.
• This research highlights the potential of advanced shared control for improving powered wheelchair usability.