Adaptive shared control strategies based in the Bayesian recursive technique in an intelligent wheelchair

Area of Science:

• Robotics
• Artificial Intelligence
• Human-Computer Interaction

Background:

• Intelligent wheelchairs require robust obstacle avoidance for user safety and independence.
• Existing shared control methods can be further optimized for diverse environmental conditions.

Purpose of the Study:

• To present an adaptive shared control method for intelligent wheelchairs.
• To enhance obstacle avoidance capabilities for users with disabilities.
• To improve overall system performance in various environments.

Main Methods:

• Developed an adaptive shared control method utilizing Bayesian recursive techniques.
• Implemented three autonomous tasks for different environmental scenarios.
• Integrated laser range finder data with user intentions for task selection.
• Estimated task evidence using Bayesian recursive techniques during wheelchair movement.

Main Results:

• The adaptive shared control method demonstrated significant performance improvements.
• The system effectively selected the most suitable autonomous task based on environmental context and user intent.
• Experimental results surpassed previously reported shared control methods.

Conclusions:

• The proposed Bayesian-based adaptive shared control method enhances intelligent wheelchair navigation.
• The system offers improved obstacle avoidance and user assistance in diverse settings.
• This approach represents a significant advancement in assistive robotic technology.