LiDAR enables this intelligent, all-terrain robot wheelchair

Partial or shared autonomy, in which the user retains most of the control, represents an unsung revolution in our relationship with machines.

In spite of advances in wearable robot technology, robots have had a hard time displacing wheelchairs as the dominant technology to help those living with physical disabilities navigate the world. Wheelchairs are too reliable and widespread, and they cost far less than sophisticated robot suits, which still have a hard time navigating uneven terrain.

But wheelchairs have limitations. They aren't great in tight spaces or difficult-to-navigate terrain. They can't go upstairs.

Also: The 10 skills you need to land a job working on autonomous vehicles TechRepublic

Now an intelligent, all-terrain wheelchair equipped with much of the same sensor and guidance technology as self-driving cars is providing a stopgap between state of the art wearables and old school wheelchairs. The project is the result of research being conducted in the lab of Ming Liu, IEEE Member and Assistant Professor of Electronic and Computer Engineering at Hong Kong University of Science and Technology.

"Most existing electric wheelchairs do not really include navigation systems," Liu told me in a recent interview. "Our project enables the robotic wheelchair to navigate in a very complex environment and to automatically move around according to the 3-D information it perceives."

The wheelchair uses a LiDAR sensor, the average cost of which has plummeted in recent years.

The robot's autonomy is meant to assist the human user by making crucial decisions in certain difficult situations. The intelligent wheelchair is an example of partial or shared autonomy, in which human pilots guide a system toward a goal but rely on the machine to make crucial decisions related to executing the goal.

"When the robot moves around, it creates a 3-D representation of the environment, and then the robot will analyze this map so that it knows where the difficult parts are and how to transform itself so it can automatically move from point A to point B," says Liu.

With the help of its tank-like treads, the wheelchair can transform to climb stairs.

Also: Is the race to autonomy another race to the bottom? CNET

"This benefits end users greatly in that they're able to move freely in areas without lifts," says Liu.

The wheelchair system is currently not commercialized. Liu believes there are still hurdles, although he believes such a system can be brought to market.

"All autonomous systems require the combination of perception, decision making, planning and control. Each aspect is in need of state-of-the-art research output and efforts. From the commercialization perspective, we still need time to educate consumers, to help them better understand the safety of wheelchair robots and to educate them on how to use these robots."

Liu believes with falling sensor prices and the increasing pace of research into autonomous systems, healthcare robots will be in much wider use in the next five to ten years.

Previous and related coverage:

BMW partners Innoviz for solid-state LiDAR autonomous vehicles

BMW is going with a solid-state LiDAR system for its self-driving vehicles, which it plans to put into production by 2021.

Drone captures LIDAR map of EastLink for autonomous car trials

The Telstra-provided aerial drone captured a LIDAR map of Melbourne's EastLink tunnel in preparation for trials of fully autonomous cars.