Background

This application was made as a part of the Capstone Applied AI project (TI3150), a course in the minor Engineering with AI at the University of Technology Delft. It was created by Alan Roukema, Gijs Volkers, Skip Doorn, Sophie Vlot and Ian Tiemann.

This is a project description from projectforum.tudelft.nl:

“Data driven design offers great opportunities for making sense of what products and objects do ‘in the wild’, using cheap sensing to generate new kinds of design insight in real time. Sensor data can be hard to make sense of, but machine learning can help: classification, novelty detection, dimensionality reduction can all help to make engaging with datastreams more meaningful, and are part of relating low level data to human experience. This project looks at collecting data from an object with sensors on, and creating ML classifiers that can decide what kind of action/activity is being performed, to support designers’ understanding.”

We would firstly like to thank our clients Jacky Bourgeois, Dave Murray-Rust and Kostas Tsiakas for their support and helpful feedback during the project. And many thanks to our TA, Matej Havelka, for technical advice and always thinking along.