Help for standing up, walking, sitting down

A robot assistant recognizes the users’ state through just a few sensors, which also calculates the user’s center of gravity.

When the robot user leans forward, the robot detects it and assists in standing up by elevating the armrests.
PHOTOS COURTESY OF TOYOHASHI UNIVERSITY OF TECHNOLOGY

A research team led by Mizuki Takeda, an assistant professor at the Department of Mechanical Engineering in Toyohashi University of Technology, developed a robot that can assist standing up, walking, and sitting down. By estimating the user’s state using a small number of sensors, they calculated candidates’ center of gravity. Knowing the user state, such as trying to stand up or sit down, makes it possible to assist and prevent accidents by detecting irregularities, such as the user almost falling over.

The support robot at the sitting level (left) and standing level (right).

Conventional robots have assisted in helping people stand or walk, but since standing up, walking, and sitting down are done in sequence, it’s desirable to have a single robot assist with them all. Large-scale systems are often constructed to produce great force or recognize the user’s state in detail, but systems for household use should ideally be small and have few sensors. Having all movements performed centering on the robot may lead to dissatisfaction as it promotes muscular weakening, and the user feels they’re not in charge of their own movements. Because of this, researchers developed a robot putting the user in the center of the movement, recognizing the user’s situation, and adding the missing force.

The research team focused on center of gravity as key to estimating the robot user’s state. Center of gravity is an effective indicator of a person’s state, but an accurate center of gravity can’t be known with few sensors, so they developed a technique for calculating the range where center of gravity candidates can be found by considering the motion range of joints. The robot uses these center of gravity candidates to estimate the user’s state and detect irregularities through a type of machine learning (ML) called support vector machine and then moves according to that information.

The research team believes estimating people’s states using center of gravity candidates can be applied in healthcare robots of various shapes and uses. They want to clarify what movements robots should perform and how they should communicate with people to become easier for the elderly to use.

Department of Mechanical Engineering in Toyohashi University of Technology: https://www.tut.ac.jp

March 2022
Explore the March 2022 Issue

Check out more from this issue and find your next story to read.