What is the Connection Engine?
The Congruence Engine (later renamed Connection Engine) is a three-year research project funded by the Arts and Humanities Research Council in efforts to digitally connect industrial heritage collections across the UK. Using advanced digital tools such as AI, it links museum objects, archives, photographs, and more, enabling historians and curators to explore industrial history beyond traditional constraints. The project functions as a “social machine,” combining human expertise with digital methods like transcribing oral history, geolocating photos, and large language models.
We partnered with the Science Museum to create an exhibition that showcases the power of the Congruence Engine (later Connection Engine for accessibility) in researchers’ work connecting and revealing Bradford’s scattered industrial history. Our primary goal was to demonstrate how advanced digital technologies are becoming an essential tool for researchers in uncovering historical narratives. The exhibit was meant to make complex digital processes accessible and relevant so it could be accessible not only to our target audience of history buffs and local heritage but for all museum-goers. We wanted to inspire visitors of the potential of using these technologies to expand research by encouraging public participation in historical exploration through real-life Bradford case studies.
Setting up our workspace
The exhibit would primarily be a large interactive touchscreen table paired with a large digital display that would be mounted above the table in efforts to attract visitors and spark some intrigue in the exhibit. We were given a wide breadth of stories that researchers had unearthed about Bradford’s history using the Congruence Engine. We also had a few preliminary datasets and the go-ahead to request new datasets and information that would be deemed useful for the exhibit to be successful. We were also given a few user learning objectives to guide the design:
- Feel excited by the new ways that digital technology can help us experience our history, especially local stories from Bradford.
- Understand that advanced digital technologies and AI are being applied to museum and archival research.
- Be surprised and stimulated by the stories uncovered by the Congruence Engine project and researchers, working as part of a ‘social machine’.
- Be inspired by the potential of creating a connected National Collection.
In order to fulfill the learning objectives, we first needed to understand our target users. While we aim to create an exhibit that is interesting to everyone, we targeted users that would have a natural inclination for the subject material such as history buffs or community-engaged folks. While we felt these museum visitors would have a natural interest in the subject matter of Bradford’s industrial past, we knew they may also come in with hesitations or opinions about AI’s use in research.
Map as the focal point
We decided that the best way to guide users through this experience was to show how advanced digital technologies were revolutionizing researchers’ work. We wanted Bradford to be the focal point to help orient users around the stories related to Bradford’s past. Our main dataset featured stories with primary locations around the city, so a map felt like the most natural entry point for the exhibit. Exploring a map on a large touchscreen table is also typically a satisfying interaction for users as it is both familiar and intuitive. We added large zoom buttons and the ability to toggle historic map background layers to urge users to dive in and start interacting. We hoped that exploring the map would result in a user tapping a large story photo button and diving into the data.
When a user selects a story, their view takes them to all the data that is connected to the story. From here users can explore the data freely in the data buckets on the lefthand side. We wanted users to feel as though they were immersed in the archives and connecting the puzzle pieces of the story. Each data bucket is tied to a short paragraph explanation on the righthand panel to provide insight into the data. If users preferred, they could more linearly step through the story beats with buttons at the bottom of the righthand panel which would open the relevant data buckets. From here, the challenge for us would be how best to highlight the advanced digital techniques that historians used to find and connect the data presented in the data buckets.
Defining the AI narrative
Many people approach AI with skepticism, often concerned about its potential to oversimplify, distort, or even replace the complexity of human knowledge and manpower. The Connection Engine offered a compelling counterpoint, demonstrating how AI can be a powerful tool for historical research rather than a replacement for researchers. While we knew this to be true, it was important for us to consider how many museum-goers might be skeptics about the merits of the project before even engaging with the thesis. We struggled with how to posture advanced digital technologies’ role in this work as we kept finding in our initial user testing that many people were missing the learning objectives completely or had a natural disdain for AI.
Many visitors were either so fixated on the stories themselves that they didn’t realize how the stories had been collected, or they were worried that AI was replacing the researchers that were doing this work. We ultimately decided that the best way to convince museum goers of the work was to keep the primary focus on Bradford while weaving in story-specific digital techniques as motifs throughout the experience. We added icons and user-friendly language for each of the digital technologies as buttons throughout the experience. Users could open more information about each technique at any point with one of these buttons, or keep exploring stories with the subtle advanced digital technologies callouts.
Weaving in the techniques
When a user chose to learn more about a technique, their view would change to show information on what the technique was and how historians used it, as well as interact with a simulation to convey vaguely how the technique worked. The simulations were also a way of “gamifying” a relatively complicated topic. For geolocating photos, for example, users have a few images that they are meant to drag onto a map by recognizing the location from the photo or matching the coordinates.
Historians also trained large language models to help them connect information across galleries about the industrial era in Bradford. They could ask the model questions to add insight to their research. We simulated this approach by creating a series of question buttons that users can tap and see an animated ellipse transform into a pre-determined response.
We added small design elements like sound effects and animated videos to help build out the effect of the simulation and keep users engaged. For example, historians used optical character recognition to transform an image of text into lines of text. For this simulation, users are presented with a fuzzy image of text overlaid with a camera shutter design. When they tap the image, a shutter sound effect plays and the image is replaced with crisp lines of text.
Curating the vibe
The exhibit was not complete without a three meter tv screen display above the table designed to attract users to the table and spark intrigue. We went through several iterations of ideas but knew that we wanted the large digital display to feel exciting and loosely call out both the techniques and people-power that power The Connection Engine. We wanted users to feel like they were immersed in the data and like these advanced digital technologies were a part of that feeling. However, we did not want users to feel like they needed to look back and forth between the table and large digital display in their experience. We created a design that brought in images from all the stories in a scrolling format. Interspersed between the photos were icons and definitions of the digital technologies to add to the atmosphere.
We ensured the visual design of the large digital display was in line with the interactive touchscreen to create a seamless experience. The visual design was meant to feel immersive, futuristic, and minimal to convey a sense of exploring the galleries and digital technologies without detracting from the imagery or other pieces of data. We achieved this look by using large, clean, accessible text throughout and a simplified, bright color palette. We leaned on a dark mode design but brought in lighter elements such as the map to ensure there was enough contrast with the story photos and similarities with the lighter historical map layers.
Over the years we’ve always had a blast designing large touchscreen interactive experiences for museums, like the Exploratorium for example which you can read about on our website, and this was no exception. While navigating preconceptions about AI was tricky, we love a good design challenge and found the Science Museum to be excellent partners in tackling the problem. If you have any museum-grade exhibition needs or just a tricky design problem, drop us a line!