Initial preparatory work

Due to the amount of uncatalogued material held offsite, prior to any work commencing, the first step was to make sure as many Deepstore barcodes, collection reference numbers, accession numbers and box numbers were matched up as possible. The bulk of this work was undertaken in December 2019 and identified discrepancies between our collections information and our storage, movement and location control information. For instance:

• Many Deepstore barcodes had no reference associated with them.

• Accession numbers were often not differentiated, using only the collection reference as an identifier.

• Previous accessioning practices meant individual accessions could be as small as a single file and stored within other accessions for that collection, making them very difficult to locate.

• Previous accessioning practices meant accessions containing multiple collections could be recorded multiple times to reflect both the individual and collective accession, confusing catalogue and extent status.

• Extents varied between the collection file, CALM and the Deepstore inventory. Reboxing, part-cataloguing and weeding processes have not been consistently documented, making it difficult to confidently ascertain the true extents of accessions and collections.

• Data has been entered in different places in different forms, even within the same format i.e. CALM accession records, box lists, collection surveys. It requires a great deal of manual work to untangle.

Following this data gathering exercise, a practical two-week long pilot sprint took place in February/March 2020 to understand how to process different kinds of boxes and the time taken. The pilot included a selection of unidentified barcodes, a collection with over 10 small accessions, and a number of other mid-sized collections (25-50 boxes).

The basic Deepstore inventory data was further improved during 2020 by LE&E (Library Experience & Engagement) to connect previously unknown references to barcodes.