Evaluating collection management systems for interdisciplinary natural history collections

See also the recording available at https://vimeo.com/174716383. ABSTRACT. A collection management system (CMS) enables collection managers to do their jobs more efficiently and with more impactful results. In this case study, the Chicago Academy of Sciences / Peggy Notebaert Nature Museum (CAS/PNNM) shares its experience choosing between four CMS options: EMu, Specify, Arctos, and CollectionSpace. Each was evaluated on its functionality – including its ability to accurately represent our diverse collections and to enable efficient workflows – its provision for data discoverability, and its feasibility for implementation. Because the technical specificities of software inherently have a short lifespan, this case study offers evaluation through a broader framework of information science that should continue to be relevant even as the CMS options evolve. INTRODUCTION & METHODS. Many museums have faced the task of acquiring a new collections management system (CMS) that is relational database software designed to specifically address the needs of collections. Investing in the right CMS should enable collection managers to do their jobs more efficiently and with more impactful results. The task of implementing a new CMS, however, is daunting, in large part because technology changes so rapidly and because the process requires a major commitment from the institution in terms of staff time, hardware upgrades, and software costs. The Chicago Academy of Sciences is the oldest science museum in Chicago, with biological and cultural collections dating back to the 1830s. A digital inventory project from 2008-2013 resulted in 61 Excel spreadsheets containing upwards of 285,000 specimen label data records. However, maintaining data in so many different spreadsheets was unsustainable. In order to provide sufficient quality control and security of the data, and to enable more effective management of the collections, CAS/PNNM decided to invest in a collections management system. Initially, CAS/PNNM Collections staff cast as wide a net as possible in considering available collection management systems, particularly those specific to natural history collections. The available literature regarding CMS options was sparse and often the specifics of the software were found to be out of date. iDigBio and the Canadian Heritage Information Network have made available summaries about some of the CMS options, and two unpublished master’s theses also produced valuable comparisons of different systems (Sully 2006 and Carpinone et al. 2012). Coetzer et al. (2012) provide details about one CMS option, Specify, while Lathrop, McDaniel, & Ritchey (2010) describe another, PastPerfect. We also requested anecdotal information from the museum community through the NHColl listserve, which responded with helpful but scattered opinions. Through this scan of existing literature and resources, we were able to narrow down our initial broad sweep of CMS options to those that would be most suitable for our collection: EMu, Specify, Arctos, or CollectionSpace. CMS options that we briefly evaluated but excluded from the remainder of this process were: Adlib, ArchivesSpace, ARGUS, Biota, CollectiveAccess, Mimsy XG, PastPerfect, Re:discovery, Symbiota, The Museum System (TMS), and Vernon CMS. We also declined to assess developing a custom database. RESULTS. Evaluating Functionality. We examined the ability of each CMS to accurately represent collection objects and to enable efficient workflows, including (1) essential collection management transactions, (2) data cleaning and batch editing, (3) georeferencing, (4) digital asset management, and (5) content management. Our evaluations were based on thought exercises and the demo versions of software that we had access to. All four CMS options were sufficiently functional, although the specific functionality offered by each varied more than maybe expected, and no single solution has managed to incorporate every great tool (see Figure 1). Evaluating Data Discoverability. All four CMS options provide the core search and browse functionality expected of a relational database. EMu, Arctos, and CollectionSpace have the ability to search across all collection types within the CMS. Specify provides full-featured searching within a single collection, but runs into barriers trying to draw information frommultiple collection types. In addition to internal searching, online data discoverability is an increasingly relevant aspect of biodiversity data discoverability. While Specify and EMu offer built-in pathways to online data publishing, and CollectionSpace can be set up to do so, Arctos makes publishing to online data aggregators as easy as possible. Evaluating Feasibility of Implementation. It was essential to round out our evaluation of CMS options by considering both the direct costs of software licensing and support, and the indirect costs of new hardware, infrastructure, and staff time. Specific figures from our own institution’s direct and indirect cost calculations are provided for reference. Our examination determined that all CMS options cost more than they appear to upfront – even “free” options such as Specify require considerable staff time for set-up and maintenance. Arctos provides perhaps the greatest indirect cost benefit to small, multi-disciplinary collections by allowing members to take advantage of essential curatorial decisions already made by the Arctos consortium body’s expertise, as well as by centralizing technical support for the CMS. DISCUSSION & CONCLUSIONS. One of our major deciding factors at CAS/PNNM was the ability for a CMS to handle all of our collections – from anthropology to ornithology to audiovisual – within a single database. Specify, although certainly a viable and affordable option for more taxon-specific collections, did not make sense for the diversity of ourcollections. Additionally, staff capacity was an important concern. We were very attracted by the process-centric approach of CollectionSpace, as well as its user interface design and obviously dedicated core community, but the time investment required would be greater than feasible for us. Likewise, after considering the learning curve and design time commitment required by EMu, we felt that it too would require more staff time than we have available. Arctos would stillrequire a significant time commitment; however, the implementation process is significantly streamlined as the interface layout and functionality have already been established. This means we would be able to progress immediately into preparing our data for migration and learning to use the new software. In addition, our IT staff wouldn’t be overburdened with the installation and implementation process, nor with future CMS technical needs. There are many solutions for museums to manage their collections data and ultimately, one size does not fit all. It is crucial for each institution to thoroughly consider the scope of their collection and what they want from a CMS. For CAS/PNNM, our collections represent a wide range of disciplines, and we wanted to maintain and expand on the connections bridging them. We determined that for our situation, Arctos would provide the easiest path to getting quality, standardized data into a relational database for all of our collections, and to making this data accessible both in-house and online. REFERENCES Canadian Heritage Information Network. 2012. Collections management software criteria checklist. http://www.rcip-chin.gc.ca/carrefour-du-savoir-knowledge-exchange/criteres_gestion_collections_management_checklist-eng.jsp (10 April2015). Carpinone, E.C. 2010. Museum collections management systems: One size does not fit all (unpublished master’s thesis).Seton Hall University, South Orange, New Jersey. Coetzer, W., O. Gon, M. Hamer, F. Parker-Allie. 2012. A new era for specimen databases and biodiversity informationmanagement in South Africa. Biodiversity Informatics (8): 1-11. iDigBio. 2011. Biological collections databases, tools, and data publication portals. https://www.idigbio.org/content/biological-collections-databases (10 April 2015). Lathrop, S., S.L. McDaniel, N. Richey. 2010. Exposing ourselves: A case study in collection management softwareimplementation. The Southeastern Librarian 58(1): 3-8. Sully, P. 2006. Inventory, access, interpretation: The evolution of museum collection management software (unpublishedmaster’s thesis). John F. Kennedy University, Pleasant Hill, California.