Reimagining Mobility: CMU Helps Develop Next-Generation Robotic Wheelchairs
Carnegie Mellon University researchers are helping lead a major effort to redesign the powered wheelchair from the ground up—this time with robotics at the core, not as an accessory.
As part of a $41.5 million ARPA-H funded initiative, CMU’s Robotic Caregiving and Human Interaction Lab will integrate a robotic arm directly into the wheelchair’s architecture, enabling users to grasp objects, open doors, pick items off shelves, and perform everyday tasks at home and in public spaces.
“The idea is to create the future of wheelchairs,” said Zackory Erickson, assistant professor in CMU’s Robotics Institute. “Instead of attaching a robotic arm after the fact, the chair will be designed around the arm, making it part of how users navigate and interact with the world.”
The project, known as the Robotic Assisted Mobility and Manipulation Platform (RAMMP), is led by the University of Pittsburgh and includes Northeastern University, Cornell University, Purdue University, and industry partners ATDev, Kinova Robotics, and LUCI Mobility.
The redesigned wheelchair will combine:
Advanced robotics and shared-control autonomy
A new operating system for real-time awareness
Digital twin simulation for training and testing
360-degree sensing to prevent tipping, collisions, or unsafe movements
Most importantly, the CMU team is focused on making this robotic assistance easy, intuitive, and fast. Current robotic arms mounted on wheelchairs often require slow, painstaking precision to pick up everyday items. The CMU team will develop learning-based algorithms that help the arm automatically align with objects, grip securely, and perform fine-movement tasks.
“Precision matters,” Erickson said. “If we can automate part of the movement and share control between the user and the arm, we can make everyday actions smooth instead of tiring.”
The team will also equip the system with cameras, tactile sensors, and haptic feedback—allowing users to feel and interact more naturally with their environment.
Once developed, the system will be tested directly with wheelchair users performing real-world tasks such as grocery shopping and household chores. User feedback will shape the final design.
The work builds on the lab’s history of robotics for caregiving, including systems for dressing assistance, feeding, and household support, as well as the head-worn HAT teleoperation interface previously tested in-home with a quadriplegic user.
The project seeks to expand independence and agency for millions of people with mobility disabilities, including veterans, and to shift assistive devices from passive support to active, intelligent partners in daily life.
