The following is a guest post by Matt Martin, Chief of the Collections Management Division. Matt discusses the process of designing and implementing a new inventory management system for offsite storage.
Over the years, the Collections Management Division has posted many blogs that reference our offsite collections. This includes our most recent post about the ten millionth item sent to offsite storage, and a previous post about planning offsite storage shelving. While facility planning and collection processing are critical factors in managing an offsite storage program, nothing works if there is not an Inventory Management System keeping track of it all.
Last summer, we posted a brief introduction to an initiative within the Collections Management Division to develop a new Inventory Management System, called the Finding EnvironmenT for Collected Holdings (FETCH).
We are pleased to announce that as of Spring 2025, FETCH is fully implemented at the Library. As you can imagine, developing a new system from scratch is not an easy process as it requires significant input from users, IT security staff, developers, and many other stakeholders.
How does such an implementation succeed? The key for the Collections Management Division was to start early. While system development was underway, in January 2025 we began a training program with our staff across three separate locations (Cabin Branch, Fort Meade, and Capitol Hill). Our previous Inventory Management System was DOS (Disk Operating System) based, which required keyboard commands for navigation. FETCH, however, is a browser based application and was a significant leap for some of our staff. For many, though, FETCH looked and behaved like many other systems they already use on a daily basis. As Michael Davis, one of our offsite material handlers put it: “It was like moving from a land line phone to a smart phone.”
As a busy division, our training program had to fit into the daily activities of our staff. Working with their supervisors, we carved out a few hours each week to focus solely on training. Training was happening as we were finishing the development of the system, which gave us an opportunity to get feedback from staff that could be turned right over to our developers for improvements to be made. For example, the staff suggested that the pick list for requests automatically move scanned items to the bottom of the list in order to ensure that the items that still need to be retrieved were at the top. They also had input for additional data to add to reports, buttons that should change locations, and much more.
Apart from training staff and working to develop the system, we also worked to prepare the data from our previous system, LAS (Library Archival System), for migration into FETCH. As you can imagine, over the past 25 years of use many errors had built up in our previous system. FETCH also contained additional data fields not found in LAS, so all of the records needed to be updated to include these new fields. In all, we migrated more than 9.9 million item records, 600K tray records, and 100K location records. Additionally, the migration tool tracked errors that were captured during the process so that they could be addressed after the migration.
In order to reduce the impact on patrons, the transition to staff use of FETCH was done on a Friday to Monday. Thursday evening after operations were complete for the day, a large data export was performed in LAS. Friday morning, we completed the addition of extra data fields that we would now be tracking in FETCH. By Friday afternoon, we had transitioned the data set to our developer team who began the full upload into FETCH. This process lasted until Sunday morning and by that afternoon our team was able to start quality checking the data within FETCH. When our staff returned to the office Monday morning, we were prepared for system-wide testing of all of our processes. Groups of staff were tasked with completing specific actions within the first few hours so we could confirm all aspects of the system were working as expected. The transition was relatively smooth, although we did encounter a few hurdles. One such obstacle was that the memory used for running simultaneous large tasks caused the system to perform slower than anticipated. The Library’s Chief Information Officer teams and the developers worked together to identify the issues and correct the problems, and by the end of the day we were operating at full capacity.
On Tuesday April 29th, our staff reported to work and business resumed as normal. More than two years of requirement definitions, software development, system testing, and training had culminated in the full implementation of our new Inventory Management System. It has now been almost 6 months since our go live date. In that time, we have added more than 100,000 items to our offsite storage facilities and fulfilled more than 24,000 requests using FETCH.
We are excited that the full software has been posted to GitHub with an open source license making it available for anyone to download and try out. The repositories can be found here: https://github.com/search?q=org%3ALibraryOfCongress+fetch&type=repositories
For more information on how the software works here is a link to our demo video.