It should come as little surprise that Stanford is playing a large role in the rapid progress towards fully autonomous vehicles. Research data and video recorded by John Kegelman, Lene Harbott, Chris Gerdes and others in the Dynamic Design Lab are now deposited and streamable from the SDR. These data are useful in a variety of ways, such as to inform self-driving cars that can respond to changing conditions like an expert driver handling a race car on a track.
A guest post from one of our Road & Track project archivists, Beaudry Allen:
There is always something unexpected to find when processing a collection. You do not have to be a car aficionado or even know the first thing about cars to at least have a slight remembrance of the car in film Chitty Chitty Bang Bang. While the memories evoked by the car may be its ability to fly or float in water, the car was based on the legendary Brooklands cars of Count Louis Vorow Zborowski. Zborowski was a famous 1920s English racing driver and automobile engineer known for building his own race cars, some of which were called “Chitty Bang Bang.” Ian Fleming was influenced by Zborowski’s engineered car and its eccentricities when he wrote the famed children’s story of the same name. When the 1968 film adaptation started, mock-ups were built in the Edwardian-style. They actually worked, but apparently in the style of the day the cars only had brakes on the rear wheels, which meant that there were no brakes if you went in reverse. So the car may not be safe for the road today - but certainly one for memory lane.
The Road & Track collection is currently being processed, but a portion of the archive is available. A preliminary guide is available here:
The Revs Digital Library website contains nearly 200,000 images of automotive history (as of January 2015). As with other digital library sites, users can search and browse to find content that interests them.
A search engine, however, is only as good as the data being searched on. Since the website will eventually have over a million images, browsing to find the images you want is not a great option and good information about the images becomes critical. Interested in Porsches? Chevy's from the 1960s? Mario Andretti, but only when he drove in the Indy 500? All of these wonderful ways to find materials are dependent on having metadata, or information about the images.
In an unassuming low-rise building on a side street in Naples, Florida sits the Revs Institute. The Institute, which is open to invited scholars and guests, houses a collection of fully restored historically significant automobiles, as well as a library containing images, books and ephemera. Since the images are carefully stored, many as negatives, a large number of them may not have been seen since they were taken. Up until now, this entire collection was housed under one roof, one large hurricane away from being damaged or lost.
We haven't yet figured out how to digitally preserve automobiles, but digitization of the large and unique image collection of the Revs Institute is underway. Working with Pixel Acuity and the Revs Institute, over 100,000 images representing over 1 terabyte of data have been digitized and preserved in the Stanford Digital Repository (SDR) in just over a year since the project began. The Revs Digital Library website, built on top of SDR by engineers in the Stanford University Libraries, allows users to search and view all the images. The website is currently available only to members of the Stanford University community.
Each image can be viewed at full resolution using the zoom and pan controls familiar to users of Google Maps. Collections can also be browsed or viewed in a slideshow format. Next we plan to engage the automotive community by adding tools to help improve the metadata, thus adding additional value to the collection.
The initial internal launch of the Revs Digital Library is an important milestone - not only does it ensure these images will be preserved for the future, it also makes them readily available to scholars in the Revs Program at Stanford for research purposes.
But the work is only beginning. Over the next three years, along with additional tools and features, another 300,000 images are expected to become available in the digital library. And the history of the automobile will continue to be preserved for future generations.
The Revs Digital Library is a project within the Digital Library Systems and Services group whose goal is to ensure access and preservation of materials from the Revs Institute and the Revs Program at Stanford.
The Revs Institute in Naples, Florida is an independent educational organization that advances the scholarly study of automotive history. The Institute houses a library with over a million items, including a large and varied collection of automotive materials such as images, research books, ephemera, and specialized documents.
The Revs Program at Stanford was established to promote a new trans-disciplinary field connecting the past, present and future of the automobile. The program aims to put the automobile at the center of the university and raise the quality of academic discourse at Stanford and beyond. The program is now producing research data and generating course materials.
Working with Pixel Acuity, the Revs Institute is currently digitizing their collection of images using specialized digital cameras. Each slide, negative or print is cleaned and imaged at a high resolution. The images and associated metadata are collected and transferred to Stanford, where they are being accessioned into the Stanford Digital Repository (SDR) using an automated pipeline.
The automated pipeline is built using the Ruby programming language and relies on a "robot" framework, also developed at Stanford, for queing up and executing specific jobs in various workflows. For example, in order to be accessioned, each image must be analyzed to ensure its integrity has not been compromised in transit (by computing MD5 checksums), web friendly derivative images need to be created (JP2), images need to be moved to the digital "stacks" and preservation core, and so on. A "robot" is designed for each specific task, and tasks are organized into ordered workflows, with appropriate dependencies. Queues are established to automatically move objects through the pipeline, wtih additional servers running copies of the robots added as needed to maintain throughput.
The Revs Digital Library, currently under development, will ensure that all of the accessioned materials from the Revs Institute, as well as the original research from the Revs Program, are indexed, preserved and made available to library patrons, researchers and the general public. By digitizing materials and making them discoverable, content that was once available to a select few becomes useful and discoverable for a wide range of researchers. The Revs Digital Library is being built on top of the Stanford Digital Repository to provide a web based platform for discovery of automotive research and images. The Digital Library is developed in Ruby on Rails using open source technologies, including Blacklight, Hydra, and Fedora Commons and will allow for metadata editing, provide community features, and tools for researchers to further utilize the data.
As of October 4, nearly 68,000 images from the Revs Institute's collections have been digitized and staged on Stanford Library servers, with 1000 images accessionined into SDR. By the end of 2012, we expect to have all 68,000 images accessioned, with a digital library website for browsing and viewing the materials.