AI Aims for Autonomous Wheelchair Navigation

2026-03-20 · Source: IEEE Spectrum · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Robotics & Autonomous Systems, Emerging Technologies & Innovation · Depth: Intermediate, short

Summary

New research from the German Research Center for Artificial Intelligence (DFKI), co-led by Christian Mandel and Serge Autexier, is exploring AI-powered smart wheelchairs designed for users with severe disabilities. Presented at the CSUN Assistive Technology Conference in Anaheim, Calif., the prototypes integrate lidar, 3D cameras, odometers, and embedded computers, supporting both semi-autonomous joystick control and fully autonomous natural language navigation via ROS2 Nav2 and SLAM. The REXASI-PRO project also tested a novel safety system combining wheelchair and room-based drone sensors. While aiming for advanced navigation and obstacle avoidance, the research acknowledges significant challenges, including cost, reliability in varied real-world conditions, and the diverse human factors involved, emphasizing a partnership between user and technology rather than full replacement.

Key takeaway

For AI scientists developing assistive mobility solutions, your focus should shift from full user replacement to designing robust human-AI partnerships. Prioritize explainable AI, ensure reliability across diverse real-world environments, and consider the high cost and varied user needs to make advanced smart wheelchair technology both effective and accessible within the next decade.

Key insights

AI-powered smart wheelchairs aim to enhance mobility for users with severe disabilities through advanced sensing and navigation.

Principles

• Prioritize user support over full replacement.
• Reliability is paramount in real-world conditions.
• Explainable AI is essential for user confidence.

Method

The REXASI-PRO project utilizes two lidars, a 3D camera, odometers, and an embedded computer, integrating ROS2 Nav2 for natural language navigation and SLAM for mapping, alongside a multi-sensor safety system.

In practice

• Integrate multi-modal alerts for obstacle detection.
• Use natural language for autonomous control.
• Combine on-device and environmental sensors.

Topics

• Smart Wheelchairs
• Assistive Technology
• Autonomous Navigation
• Sensor Fusion
• Explainable AI

Best for: AI Scientist, AI Researcher, Robotics Engineer, Research Scientist

Related on AIssential

• Achieving my vision of personal AI health assistants — Realizing global AI personal health assistants requires understanding user needs, advancing LLM capabilities, and overcoming integration and trust barriers.
• Video Friday: AI Gives Robot Hands Human-Like Dexterity — Advanced AI and robotics are enabling human-level manipulation, extreme flight, and complex autonomous inspection.
• How AI in Transportation Systems Is Transforming Modern Mobility — AI fundamentally improves transportation by optimizing efficiency, safety, and responsiveness across diverse applications.
• How AI is making assistance for the visually impaired more accessible — AI integration can enhance accessibility and affordability of assistance for the visually impaired by augmenting human services.
• The Future of Physical AI Isn’t Smarter Robots, It’s Smarter Interfaces — The future of Physical AI lies in making the human body a direct, low-latency interface for connected machines.
• An assistive robot learns to set and clear the table by observing humans — Robots can learn complex domestic tasks by observing humans and coordinating their limbs.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by IEEE Spectrum.