Mix & Match Machine Learning Toolkit
Research – Design – AI- Tangibility – Education
The Mix & Match ML toolkit allows designers to learn about and brainstorm with ML. The shapes of the tokens represent the constraints of the different types of ML and data. By placing the tokens on the sensor board, information and examples are displayed on a connected laptop.
The toolkit is published open source and we would love to see it being adopted in education and/or design practices.
Supervised by: Sara Colombo
Featured at: Design United DDW’22 (online magazine & research afternoon demo)
Design & AI symposium 2023
The Mix & Match ML toolkit allows designers to learn about and brainstorm with ML. The shapes of the tokens represent the constraints of the different types of ML and data. By placing the tokens on the sensor board, information and examples are displayed on a connected laptop.
The toolkit is published open source and we would love to see it being adopted in education and/or design practices.
Supervised by: Sara Colombo
Featured at: Design United DDW’22 (online magazine & research afternoon demo)
Design & AI symposium 2023
The Mix & Match ML toolkit was developed during the last year of my master Industrial Design at the Eindhoven University of Technology. Originating from the observation that ideating with AI/ ML is very difficult as there are so many possibilities, I started designing a toolkit that could provide just-in-time knowledge to designers in a playful manner. Leveraging the advantages of Tangible User Interfaces, I embedded the constraints of the types of ML into the shapes of the tokens. The tangible tokens are sensorized and once placed on a sensor board, information, and examples are shown on a web interface. Since designers mainly understand ML through exemplars, real application examples are included for each combination of data and ML capability.
