Book scanning technology has seen many advances in the two decades since the U-M Library began to digitize its collection. But the world has yet to see a low-cost, automatic page-turning scanner that would be within the reach of small libraries and cultural organizations seeking to preserve local collections.
Students in Mechanical Engineering (ME) 450, taught by Dan Johnson during Fall semester and Wei Lu during Winter semester, decided to take on this challenge.
ME 450 enables students to apply the knowledge acquired in their coursework by developing creative approaches to real-world engineering problems. Over two semesters, students worked in collaboration with U-M Library staff to build a linear book scanner from an open-source Google design.
The design originated with former Google engineer Dany Qumsiyeh, who sought an alternative to the high-end, high-volume scanners that Google deploys in its mass digitization work. A vacuum cleaner, an inexpensive Canon document scanner, and some sheets of aluminum made up the core of Qumsiyeh’s prototype scanner. But in a presentation at the U-M Library, he acknowledged that there were some wrinkles to work out, and that the $1500 he spent on materials could surely be reduced.
ME 450 student Ryan Snyder said, “The linear book scanner intrigued me from the start because I knew it would provide value for the university and further educational resources.” Teammate Lauren Staszel said she was particularly interested in the creative challenge of improving the design to make it a viable product.
A team of librarians including Kat Hagedorn, Meghan Musolff, Jim Ottaviani, and John Weise set out the desired modifications: the scanner should accommodate books of different sizes, operate with an acceptable noise level, provide industry-standard image quality, and dependably turn each page without damage. In order to maximize the class effort, the team narrowed the focus to making substantial improvements to the page turning mechanism. The library team provided support and attended design reviews, and Qumsiyeh tuned in remotely to answer technical questions
Students rebuilt the body of the scanner, using a slippery plastic to reduce friction and a steeper angle for the book so that the tough book spine would absorb more force than the delicate pages. They also worked on the perfect balance of vacuum suction to turn a single page – plus a sensor to indicate when a page wasn’t turned, when multiple pages were turned, or when a page had existing damage.
One of the library’s requirements was that prototype designs be released as open source. “This is one of the first officially open-source projects in the class,” said Johnson. “This type of project is very attractive because we want the students to be able to make an immediate impact and be able share their work with as many people as possible. It also ensures that future teams, and people from around the globe, can access the project files and continuously improve it; a win-win for everyone.”
The resulting prototype scanner cost approximately $1000, making it an exciting advance toward the creation of an affordable and scalable book scanner for small libraries and the cultural heritage community.
John Weise, manager of the library’s Digital Library Production Service, said, “This project has the potential to result in a scanner that will make preservation book scanning affordable for libraries of all sizes, around the world. It’s great to engage students in the process, and give them an opportunity to make such an impact. The results have been amazing so far.”