Science outreach
As an Open Labs science volunteer at The University of Manchester, Dr. Omar Costilla Reyes demonstrated the smart-carpet sensor system to the public and explained how machine learning is used to interpret the walking patterns it records. His session, “What can machine learning infer about the way you walk?,” was scheduled for July 26, 2016, the same week he presented a poster on the same subject at the EuroScience Open Forum 2016 in Manchester as a Marie Curie young researcher.
The smart carpet and gait
The smart carpet, developed in Manchester’s Sensing, Imaging and Signal Processing group, is a floor-sensor prototype that uses plastic optical fibers to record spatio-temporal information from footsteps. It was the platform at the center of his doctoral research on gait analysis for healthcare and security. Gait is the way a person walks, and machine-learning models can be trained to find regularities in the signals a footstep produces, for example to tell different walking activities apart, including walking while doing a mentally demanding task. Researchers are interested in this for security verification and for possible early markers of neurodegenerative disease. The carpet is a research tool, and a public demonstration lets visitors see how a floor can recognize the way a person walks.
Other outreach around the carpet
Open Labs was one of several outreach activities built around the system. He also demonstrated it at the university library’s DigiLab, ran a machine-learning workshop there in March 2017, served as a STEM ambassador at Manchester’s Museum of Science and Industry from 2016 to 2018, and was scheduled to show how machine learning can address healthcare problems at the university’s first community festival in June 2017.