MobileHCI '25 Adjunct: Adjunct Proceedings of the 27th International Conference on Mobile Human-Computer Interaction • 2025

ODISI: On-Device Intelligence for Smart Interactions

Bo Zhou (German Research Center for Artificial Intelligence (DFKI) and RPTU, Kaiserslautern, Germany), Passant Elagroudy (German Research Center for Artificial Intelligence (DFKI) and RPTU, Kaiserslautern, Germany), Paul Lukowicz (German Research Center for Artificial Intelligence (DFKI) and RPTU, Kaiserslautern, Germany), Daniel Geißler (German Research Center for Artificial Intelligence (DFKI), Kaiserslautern, Germany), Mengxi Liu (German Research Center for Artificial Intelligence (DFKI), Kaiserslautern, Germany), Sizhen Bian (German Research Center for Artificial Intelligence (DFKI), Kaiserslautern, Germany), David Cuartielles (Arduino and Malmö University, Malmö, Sweden), Felix Dietz (University of the Bundeswehr Munich and LMU Munich, Munich, Germany), Oliver Hein (University of the Bundeswehr Munich and LMU Munich, Munich, Germany)

Human-Centered AIembedded machine learninginteractive systemssmart textileson-device AIwearables

Highlights

  • Focuses on deploying AI models directly on microcontrollers, smartphones, and smartwatches for real-time, private, offline-capable interactions.
  • Explores smart textiles as rich input modalities, including capacitive, tactile, pressure, motion, and physiological sensing.
  • Covers end-to-end pipelines from sensing to on-device inference, emphasizing efficiency, latency, and energy use.
  • Addresses interaction design, usability, and human factors for AI-driven wearable interfaces.
  • Includes hands-on tutorials (smart interaction and on-device intelligence) and collaborative prototyping activities.
  • Discusses ethics, privacy, security, and user trust in edge AI systems.

Abstract

As interactive technology moves toward seamless and natural user experiences, there is a need for adaptive and responsive interactions that react to user behavior and environmental changes. Embedded AI enables real-time, on-device processing, reducing latency, enhancing privacy, and maintaining functionality offline. Smart textiles provide versatile input through touch, motion, and physiological sensing. The ODISI workshop explores integrating these technologies, focusing on deploying AI models on resource-constrained devices such as microcontrollers, smartphones, and smartwatches. By combining embedded AI with smart textiles, the workshop advances real-time adaptive interfaces, addressing deployment strategies, innovative sensor integration, design considerations, efficient processing pipelines, and ethical issues. Through interactive presentations, hands-on tutorials, and collaborative activities, this full-day event fosters insights and developments in AI-driven wearable and mobile interactions.

Click to read Full Paper.

Want to stay up to date with all the information?

Sign up for our newsletter!