The Squat Bot
Team: Kayla Blalack, Leo Wang, Maximus Maldonado, Lauren Marbury, Prof. Ye (Sarah) Sun
The Squat Bot is a motorized lower-limb exoskeleton for assistance standing and sitting with a specific focus on elderly patients. Developed with data from observed test subject motions and widely-accepted kinematic models, the prototype provides assistive torque to the knee and hip joints when standing and sitting. The work was accepted for publication and oral presentation at IEEE BSN 2024.
My work on the project focused on the electronics and motor control. I paired this capstone work with my thesis “Function Augmented by Beauty: Assistive Medical Devices for Patient Empowerment.” It explores the ways in which a consideration of aesthetics can add function to medical device designs. The thesis was a finalist for the Hughes Award given to the best senior theses across UVA’s School of Engineering.
The Results
The Squat Bot prototype, shown below, provides the wearer with gentle assistance and support when sitting and standing. All electrical components are stored in a backpack for full portability of the system. The project was completed in four months for under $1700, which is far below the timeframe and cost to manufacture any device on the market.
Background
Aging makes sitting and standing more difficult. As humans age, they become significantly more at risk of falling due to a loss in muscle and balance control, and also adopt more sedentary lifestyles. Elderly patients provided with an assistive device for everyday use will be healthier and at a lower risk for injury.
The only precursors to our design on the market are military suits, warehouse exoskeletons, and rehabilitation exoskeletons. They are meant for moving heavy objects and for medical research, which makes them overbuilt for daily use, expensive, and primarily experimental. They have also not yet focused on sitting and standing. There is no pre-existing solution for standing and sitting calibrated for everyday assistance.
Mechanical Design
While my focus within the group was on electronics and controls, I advised on mechanical design throughout the project and conducted the calculations necessary for motor selection and device safety.
Control Design
In order to understand the movement of each joint when sitting and standing, I consulted existing research and conducted my own video analysis. Coordinates of the joints relative to one another can be easily obtained from a simple video with repeated trials. After applying geometric transformations, I plotted the joint angles over time. These motion profiles showed four phases of motion that reflected those observed in accepted biomechanical research. They also provided precise angle amounts personalized to our test subject.
I built the electronic system to run on one Arduino Nano. The motors are controlled in parallel using the AccelStepper library. Each joint is given a target position based on the angles recorded from the video capture. The motors run simultaneously to execute standing and sitting over four phases of motion. Future development would include encoder feedback for auto-adjusting PID control, but for the initial prototype the motion profile was hardcoded into the design. In future designs, I would also replace the motion toggle switch with EMG sensors to trigger the assistance with a person’s muscle movement when attempting to sit or stand.