This position paper is for the AI for Haptics and Haptics for AI: Challenges and Opportunities workshop at CHI 2026 in Barcelona, held on Thursday 16/04/2026.

Find out more about my work at CHI 2026 in this blog post.

Abstract:

The unidirectional process of mapping gesture to synthesis parameter in digital musical instruments has remained prevalent, even in more recent designs that incorporate machine learning models to performing parameter mapping and synthesis in response gestural input. Haptic feedback provides the opportunity to implement bidirectional interaction through gestures – possibly enabling the exploration of AI mappings that subvert the conventional unidirectional flow. Additionally, machine learning provides the opportunity to implement closely-coupled haptic sensing and actuation systems for use with musical interaction, without the requirement to model complex system dynamics.

Recommended citation: Davison, Matthew. ‘Haptics for Bidirectional Interaction with AI Musical Mappings’. AI for Haptics and Haptics for AI: Challenges and Opportunities Workshop at the 2026 CHI Conference on Human Factors in Computing Systems (CHI ’26) (Barcelona, Spain), April 2026.