Data-Driven Theatre & Art with Raspberry Pi

My first “Data-Driven Theatre & Art with Raspberry Pi” workshop took place on Dec 5th 2025 at CodeBase Edinburgh, co-hosted with Clare Duffy from Civic Digits C.I.C.

The session brought together a small group of theatre-makers, artists, technologists, and people working across disciplines. The focus was on how small, affordable computing systems can sense live inputs, such as faces, gestures, movement, sound, and simple biometrics, and use those signals to influence what happens in a performance or installation setting.

The workshop grew directly out of work developed for EXPOS3D, a large-scale interactive exhibition by artist Trevor Jones shown at Custom House, Leith. That exhibition made extensive use of Raspberry Pi based systems distributed throughout the space, handling sensing, control, and audience interaction.

Having recently finished running EXPOS3D, I had access to a significant amount of Raspberry Pi hardware that had already been built, tested, and used in a live exhibition context. Rather than packing that equipment away, the workshop became a practical way to reuse it and open those systems up so others could experiment with them directly.

A few days after the workshop took place, I learned that EXPOS3D had been formally recognised in a motion at the Scottish Parliament. The motion acknowledges Trevor Jones work and the exhibition’s role in exploring the intersection of contemporary art and emerging technologies. It was a welcome marker of the wider context in which the technical work discussed during the workshop had been developed.

The session itself moved between short presentations, live demonstrations, and hands-on work. The workshop slides set out a working definition of data-driven art and theatre, gave examples of artists and productions using live data and sensing, and outlined how Raspberry Pi systems had been deployed within EXPOS3D. There was also a practical introduction to machine learning in this context, focused on detection, classification, tracking, and expression or gesture recognition rather than generative AI.

At several points we used Swish, developed by Civic Digits, to pose questions during the session. Participants responded using their own phones, allowing input to be gathered quickly without stopping the flow of activity.

Live demonstrations were run on a laptop to show person detection and segmentation, hand pose tracking, face and body pose detection, emotion estimation, and simple gesture-driven sound and visual responses. After that, participants worked directly with Raspberry Pi 5 systems equipped with cameras and pre-built scripts. These systems were configured to run locally, without reliance on cloud services, to demonstrate how they can function as standalone edge compute units embedded within artworks, props, or environments.

Alongside the practical work, time was deliberately left unstructured. At points, some people continued experimenting with the hardware while others talked amongst themselves or moved between setups.

This workshop connected directly to my exhibition Bridging Worlds, which ran from 3 to 8 January 2026 in Edinburgh. Several Raspberry Pi based systems developed through this work were installed in the gallery as part of interactive artworks and small games.