Tom Smith

Student-Built Archival Tools

Student Commentary by Tom Smith

¶ 1 Leave a comment on paragraph 1 0 When people think of Johns Hopkins University they think of medicine, and rightly so. JHU prepares people for medical school better than any college in the world. The medical attitude, however, extends beyond the bounded world of pre-meds. Hopkins often values the reductionism of the scientific method and the cleanliness of the lab over holism and real-world, big-picture thinking. In looking for classes during my sophomore year at JHU, then, I was surprised to find The Power of Place taught by Melanie Shell-Weiss. The course was a practicum, and all participants would conduct oral histories as part of the East Baltimore Oral History Project (EBOH). This was a rare chance to get out of the infamous “Hopkins Bubble.” I signed up.

¶ 2 Leave a comment on paragraph 2 0 During Power of Place I helped produce several oral histories. As interviews flowed in, however, I began to see a problem. Each interview was around two hours long and included a full transcript. From my class alone, the project would amass nearly thirty hours of audio and several hundred pages of text. Making use of those materials would be tough. As a software developer, I saw a solution. I imagined a tool that could link transcripts with audio, allow people to tag important areas within interviews, and make all the EBOH materials searchable. With support from the university, I spent the next year working with Professor Shell-Weiss to make my idea into a working web application.

¶ 3 Leave a comment on paragraph 3 0 While working on EBOH, I started visiting archives. I could find few digital images of East Baltimore, and I wanted to know why. On a trip to the Afro-American Newspapers with Moira Hinderer, I found a plethora of visual materials. Very few, however, had been digitized. The main factor was cost; resources to manually digitize the Afro collection simply were not available. Again, I saw a technical solution: I would build a robot. I proposed creating an automatic scanner that could lift and digitize the Afro’s sensitive artifacts, all for less than $500. No one at JHU thought it would happen. Luckily, Professor Shell-Weiss managed to raise some funds, and I spent a summer prototyping. The result was the Gado 1, a proof-of-concept that would later scan over 1,000 images at the Afro. The price tag was $496.

¶ 4 Leave a comment on paragraph 4 0 After graduating from JHU in May 2011, I wanted to continue refining my technology. With support from the JHU Sheridan Libraries and the Abell Foundation (a regional grantsmaker), I secured enough funding to work on my project full time. In December 2011, I completed the Gado 2, a second-generation machine that dramatically outpaces the original. Since testing began, my machine has digitized over 3,000 artifacts at the Afro, at a tiny fraction of the cost of manual scanning. In February 2012, we made these images available to the general public and launched GadoImages.com to distribute them to paying licensors. I believe Gado has the potential to dramatically alter the economics of archival digitization. I would never be at this point without that first push to step outside the classroom and start thinking in broader terms.