Accepted Tutorials

Tutorials will take place on June 24th and 25th.

Data-Enabled Design: Designing Intelligent Ecosystems through Design Plumbing and Futuring

June 24th, 15:00

Abstract

During the tutorial, the attendees will learn about Data-enabled Design (DED), an approach which uses data as creative material to design for intelligent ecosystems. After a brief introduction to the method and (industrial) cases, the participants will collect data from the conference context using personalized design probes. This tutorial offers a careful balance between applied work and DED theory. The learning outcomes focus on (1) the use of data as creative material in DED, (2) how to set up a successful data-enabled infrastructure and (3) the role of data in guiding future design directions. In addition, tutorial participants will learn about two primary roles that aid the successful implementation of DED: data plumbers and data futures designers.

Speakers

  • Renee Noortman, Eindhoven University of Technology, The Netherlands
  • Peter Lovei, Eindhoven University of Technology; Philips Experience Design, The Netherlands
  • Sujithra Raviselvam, Eindhoven University of Technology, The Netherlands
  • Mathias Funk, Eindhoven University of Technology, The Netherlands

Go to top

Does my data fit? Assessing the Compatibility between New and Existing Data

June 25th, 8:30

Abstract

With the widespread use of machine learning within interactive applications, high-quality data is key to ensuring an acceptable user experience. This applies to training data as well as runtime data used for inference. Also, with the increasing demand for training data, data synthesis is interesting, but synthetic data must be similar to real data. Software Engineers thus need methods to assess data quality. This tutorial equips participants with techniques to check whether new data aligns with existing data. This tutorial offers practical insights and techniques to address the challenges in validating synthetic data and verifying if runtime data conforms to training data. Focusing on gesture recognition, participants will learn to differentiate between data suitable for the trained gesture recognizer and data that will potentially yield wrong results and thus lead to an undesirable User Experience.

Speakers

  • Marc Hesenius, University of Duisburg-Essen, Essen, Germany
  • Marvin Bachert, University of Duisburg-Essen, Essen, Germany

Go to top