Among the various types of robots that provide physical assistance to individuals with mobility impairments, standalone robotic platforms play a unique role. They may provide stability support to enhance the human users’ confidence in walking. They may “carry” the human users forward if necessary. They may also function as a robotic companion and follow the users without physical contact. They may even serve as personal transporters to carry loads (groceries, maybe?), allowing people to better enjoy life.
3D Computer Vision-Guided Robotic Companion
Rollators are a type of wheeled assistive device that are increasingly popular among older adults. Different from traditional walkers, rollators may roll on the ground with its wheels, making them much easier to maneuver. The hand-controlled brakes may stop the rollators instantaneously and stabilize the users when needed. However, rollators are still unpowered devices that require manual operation, which may be challenging for frail individuals. The attention needed to operate rollators also increases the human users’ cognitive load.
The robotic companion developed in our research addresses these limitations with its powered propulsion and smart 3D computer vision-based robot control. A pair of motor-powered wheels provide the desired mobility and maneuverability through differential steering.
To make full use of its dynamic capability, the robotic companion is equipped with a novel computer vision system, which provides reliable and accurate measurement of human-robot relative position without physical contact. Such unique functionality is obtained with an advanced image-processing algorithm that extracts the human torso contour from the images taken from the onboard 3D camera. Our experiments demonstrated that the robotic companion is able to maintain a stable relative position with respect to the user (keeping him/her at the center) when walking in typical daily-living environments. The potential for future practical use is signficant.
Quadrupedal Human-Assistive Robotic Platform (Q-HARP)
The world we are living in is full of small obstacles, which may cause difficulty and inconvenience to mobility-impaired individuals when using wheeled assistive devices. Motivated by this challenge, we are exploring the use of legged locomotion for the propulsion of assistive robots. Our recent product from this line of research is the Q-HARP, a novel quadrupedal robot that is intended to serve as a future mobile walker for individuals with mobility problems. The robotic platform is supported with four robotic legs, each of which includes two powered joints (“hip” and “knee”) for locomotion. As legged locomotion provides a much stronger capability in adapting to challenging terrains and overcoming obstacles, the Q-HARP is anticipated to provide greater environmental adaptability for real-world applications. We have constructed two generations of Q-HARP prototypes, investigated the gait planning and control algorithms of quadrupedal walking, and conducted experimental studies on these prototypes. Human experiments are being planned to test the robot performance in real-world use scenarios.
