Autonomous wheelchair developed by researchers at MIT and in Singapore promises improved independence for the disabled
After building self-driving golf buggies and autonomous electric cars last year, the Singapore-MIT Alliance for Research and Technology (SMART) has announced the development of a self-driving wheelchair at a hospital, which is an extension of the self-driving scooter launched last year. This self-driving wheelchair is able to reduce the workload on nurses and help them focus more on patient care.
In this video, Professor Daniela Rus, SMART FM Principal Investigator and a professor of MIT’s Electrical Engineering and Computer Science department, stated that the autonomous technology developed here could be applied to the self-driving lightweight devices, such as autonomous scooters and autonomous wheelchairs. This project has been ongoing for almost a year and a half. The self-driving wheelchair aims to provide disabled people with personal mobility, increasing their ease and convenience in their everyday life. Unlike traditional or electrical wheelchairs, people do not need to focus on controlling the wheelchair during a trip. Instead, they can just hop on the self-driving wheelchair, enter the destination, and the chair will send the destination to the system developed by SMART and retrieve the most efficient path. Afterwards, it will autonomously transport people to their destinations without any further instructions.
During the trip, the wheelchair uses sensors to capture the surrounding environment, and compare what it sees with what the map shows. When the vehicle encounters obstacles and other people uncharted in the map in real time, it will react and drive around them. Dr. Rus believes this new technology would bring more flexibility and greater personal mobility to the hospitals around the world.
The following features are included in this self-driving wheelchair:
- Fully controlled by an on-board computer.
- Uses three LIDARs to detect obstacles.
- Has six wheels to stabilize the kinematics.
- Has ability to easily go through tight spaces, such as narrow ramps, hallways, and doorways.
- Has small rotation radius which ensures the ability to make 360-degree turns.
The self-driving wheelchair is not only special for hospitals, but other situations as well, such as shopping malls or amusement parks for disabled people. This could help more disabled people go out and see the world.
Since this self-driving wheelchair is an extension of the scooter, it may be interesting to take a glance of that self-driving scooter built last year.
It looks similar to the wheelchair, but more like a single-chair golf cart. Similar to the self-driving wheelchair, this scooter also has an on-board computer for control purpose and sensor to send and receive the optimized path. The system behind the wheel consists of a few algorithms to ensure its functionalities:
- Low-level control algorithm, used for the response due to sudden changes in the environment, like obstacles.
- Route-planning algorithm.
- Localization algorithm, used to determine the location of scooter on a map.
- Map-building algorithm, used to construct the map.
- Scheduling algorithm, used to arrange the fleet resources.
- Online-booking algorithm, used to schedule rides.
Therefore, besides all those properties, the wheelchair is basically an upgraded version, specially designed for narrow space, which in this scenario is the hospital.
More interestingly, SMART unified and applied the same control algorithm among all types of vehicles: wheelchairs, scooters, golf carts, and cars etc, which reduces the complexity as different systems do different things. In this case, information flow between different vehicles can be transferred easily so that people do not have to modify the algorithms due to different types of vehicles. This is exactly the ultimate goal of programming, “write once, apply everywhere.” Once this vehicle system is complete, the transportation network could spread throughout the entire country to test its performance, and maybe MIT can go to the next phase of autonomous driving. However, SMART or MIT have not released further plan to the public.
Author: Bin Liu | Editor: Zhen Gao | Localized by Synced Global Team: Xiang Chen