Automated Sleep Support

I created an adaptive bed that automatically adjusts its contour to improve spine position and comfort for a user throughout the night. When a user lies on the surface of the bed, force measurements are taken along their body and the slats are indendently adjusted to create a bed contour that is customized for the body of the user. The aggregated force measurements can be used to classify the position in which a user is sleeping. At the same time, biometric data can be taken from the user to quantify the quality of sleep that the user is experiencing.

I also modeled the bed in Simulink and MATLAB and developed a cost function to take into account different parameters for how the bed's contour should be changed. For example, large movements on individual slats at once can cause discomfort, despite it being the fastest way to change the contour of the bed. Alternatively, adjusting many slats incrementally at the same time will slowly evolve the contour of the bed in a more comfortable way. The system model helped analyze these tradeoffs and determine an optimal control algorithm given a set of preference parameters from the user.

The bed frame was designed to be modular so that all components could fit in an average 4-door sedan, which solves a large issue in transporting beds. Also, since the contour ideally fits a user, a much smaller padding can be used and still feel comfortable to a user, which reduces the environmental impact compared to a conventional mattress and box spring. Finally, the bed is designed to be affordable.

This technology can be used for people with severe back pain, bedridden patients who have a tendency to develop bed sores, or people who require spinal movement over the course of the night. It can also be useful for people interested in collecting comprehensive data about their sleeping habits.

This project was done under the advisorship of Professor Robert Riener and Dr. Francesco Crivelli to complete the requirements of my Master's thesis. I was solely responsible for all of the design, development, and analysis of both the proof of concept and the model.

Publications and Recognition