Updating the IIIFarm. What We Learned From Teaching IIIF on a Raspberry Pi Network

Wout Dillen - University of Antwerp (Belgium), Joshua Schäuble - University of Antwerp (Belgium)

Presentation type: Presentation


In 2018 the authors organized a two-and-a-half-day workshop on IIIF for the first Digital Humanities Summer School at the Center for Manuscript Genetics, University of Antwerp. While we touched on important IIIF features such as manifests, image viewers and image annotation, the focus of the workshop centered on an aspect of image interoperability that is often reserved for technicians: setting up the IIIF compliant image server. Applying the didactic concept of experiential learning, we taught our students more about image interoperability from the perspective of its physical backbone. Guided by an interactive tutorial, we built a network of fifteen IIPImage servers on as many Raspberry Pis. Once installed and running, we used these servers to distribute digital facsimiles from the draft manuscript of Mary Shelley’s Frankenstein recently released in IIIF by the Bodleian Digital Library.

The students were divided into small groups. In a first team effort the Raspberry Pi network was set up with the aid of a gigabit router and further networking hardware. Each team then installed and configured their own device into a locally accessible IIPImage server. For the purpose of this workshop, each team represented a fictional cultural heritage institute that held the rights for three high resolution images of the Frankenstein collection. Only by transforming their images into Pyramid TIFFs, distributing them via their own image server, and collecting the images of other teams in self-written manifests, could “the monster” (i.e.: the entire collection) be assembled. Finally, all the teams’ manifests were loaded in a local Mirador Viewer instance on another Raspberry Pi. Guided by the tutors, every step of the workshop – from the initial setup and the image server configuration to the compilation and publication of their manifests – had to be executed individually by the team members, giving them plenty of opportunities to make mistakes and collaborate in order to fix them. The same learning mechanisms applied to us as instructors: we refined this experimental course throughout the two and a half day period by incorporating the students’ input and fixing bugs as we went along.

In this paper we want to focus on our evaluation of the student feedback, and on the process of refining the developed concept for a second, updated version of the workshop. Following the “learning by doing” approach, computer skills that had been condensed into theoretical introductory sessions will now be embedded directly into hands-on exercises in the course. Additionally, we want to extend the workshop and address the specific needs of digital scholarly editors and digital archives to process and publish images via IIIF. Finally, we will introduce a publicly available tutorial of the workshop that enables other lecturers to reproduce the workshop and adapt it to their needs, and look to a future where the workshop may be taught at different institutions simultaneously – thus making IIIF’s vision for ‘International Image Interoperability’ even more tangible for the lecturers and students who utilize it.


  • Discovering IIIF resources,
  • Audio/Visual use cases for IIIF,
  • Interoperability in IIIF contexts


  • digital pedagogy,
  • raspberry pi,
  • IIIF workshop