In this resource Kevin Bacon, Digital Development Officer for the Royal Pavilion and Museums, Brighton and Hove, discusses implementing a Museum DAMS that’s fit for purpose.
As Royal Pavilion and Museum’s digital lead my work covers a variety of areas, from managing our website to information management, but digital asset management presents some of the most pressing and complex challenges I currently face.
I must stress from the outset that I am not an expert in this area. I have been managing our digital asset management system (DAM) for several years, but much of my work now is unashamedly informed by my past mistakes. This resource is written as part of an ongoing case study into the practical use of a museum DAM, and while I hope they may provide some suggestions for good practice, I imagine that others may be able to contribute their own advice. Such advice will be very welcome.
Background to the Museums
The Royal Pavilion and Museums (RPM) is the local authority museum service run by Brighton and Hove City Council. We have been awarded Major Partner Museum status by Arts Council England, and receive over 500,000 visitors a year. We operate five sites: the Royal Pavilion, Brighton Museum, Hove Museum, the Booth Museum of Natural History, and Preston Manor.
Our collections are even more eclectic than our buildings and reflect the ambitions of our Victorian founders. Three of our collections are Designated (Natural History, Decorative Art, and World Art), and others include Local History, Fine Art, and Film and Media. While this diversity makes for rich opportunities, a Victorian treasure box presents plenty of problems for the 21st century.
At the time of writing, RPM has to adapt to enormous change; a proposal is currently under discussion for RPM to become a trust. Although we currently generate almost 60% of our own income, declining public funding means that we need to rapidly develop new ways of operating, both to increase income and cut costs.
All this is not merely background; these changes in governance and process have direct implications for the way we handle our digital assets. The potential move to trust status means that we may need to move away from the Council’s IT infrastructure, including the servers we currently use. It also has legal implications for some of our assets: some of our collections, such as many of our photographs, were produced by the local Council in its various guises. Where previously we could scan this material without concern for copyright, this material could suddenly belong to another legal entity.
The biggest impact on our DAM, however, comes from the need to work smarter, faster and with more agility than before. Our existing DAM was established in 2010, and our needs as an organisation have changed rapidly since then.
Our DAM today – Size does matter
We currently use Asset Bank as a DAM. We run two versions of the software. One is a public facing version hosted by Asset Bank, for which we pay a monthly fee. The other resides on an internal server that was purchased outright several years ago, and is only accessible by computers on the RPM network.
This rather awkward set up reflects the differing needs that were identified for our DAM in 2010. The public facing version was intended to offer images for sale, and act as a press library for registered users. It was originally intended to present a selected subset of assets from those available internally.
Although there was an ambition for the two systems to be integrated this was never implemented. The main reason for this was that many of our digital images are stored as large TIFF files. While this may, arguably, be a good lossless format for preserving the master image, other organisations and people seldom need a bulky TIFF file – a good quality JPEG at the same resolution is often about a quarter of the size with no real loss of quality.
Size is the key factor here. Our external version of Asset Bank can currently hold 50GB of data while our present server holds 1.5 TB. The external version sits on a cloud set up and we pay per month for the amount of data we store. It’s scaleable but increases in capacity come with a metred cost.
By contrast the internal server is owned by RPM and managed through the Council. However, this will be decommissioned at some point in the next 12 months as part of the Council’s move to a new data centre and, given the likely need to unbundle much of our ICT if we become and independent trust, migration to the new data centre is unlikely to be a long-term option.
On top of this the quantities of data I’ve reported here are only those known assets in these systems. Many more are held on external hard drives, USB sticks and cameras across the organisation. My current estimate is that we could have close to 50TB of digital assets that need to be dealt with and this figure will only grow due to continued digitisation and photography of our collections, but also greater use of video and 3D digitisation, both of which create much larger assets.
New needs for new times
The problems caused by this sheer quantity of data may be difficult to solve but they are at least easy to understand. It’s essentially an old problem in a digital guise. But as any collections manager will attest, problems of storage are always accompanied by problems of preservation and findability. In terms of our business needs, findability is the more pressing problem.
That may seem odd. The fluidity of digital data, and the potential for metadata to establish meaningful connections through search and facilitated browsing, would seem to suggest ready solutions. But findability is about matching human needs, and any metadata structure is only as good as it can match those needs. Identifying and facilitating those needs is a knotty problem, particularly when an asset can potentially be used by both staff and the public, and may originate from any part of RPM, having been produced in a variety of contexts.
It’s worth remarking on how rapidly the needs base for our digital asset management has changed in the last five years. A big change has come from our shift away from commercial image licensing towards encouraging greater re-use of our assets through open licensing.
As of 1 April 2015 all of our published digitised collection data and images are now available for re-use under a Creative Commons licence. As a result, the e-commerce functionality of our externally hosted DAM has been removed and its function has shifted from ailing commercial picture library to a more open asset repository.
The press library functionality remains for registered users but it’s now used in a very different way. Earlier this year we appointed our first Press and PR Officer who is now able to actively approach journalists. The old system was essentially passive and relied on press enquirers to know of, or be directed towards, the relevant assets. Now our Press Officer needs to be able to rapidly send selected hi-resolution assets to journalists, so I have been working with her on improving the workflows around this.
We now work very differently as individuals and as an organisation. For colleagues working outside of the office, an internal DAM behind a firewall is little use, and one of my Learning colleagues has remarked on how helpful it is to use the external DAM when preparing a teaching session the night before it is delivered.
As we increasingly work with other organisations and people, through partnership and co-production projects, an internal firewalled DAM makes little sense. I recently conducted a survey amongst some of my colleagues on their use of the DAM and the most striking finding was that four times as many people expressed a need to share assets with people outside of the organisation than within it.
These various needs were not factored into the original implementation of our DAM and to some extent they didn’t exist before. With the spread of greater digital literacy across RPM my colleagues are using digital assets more and more and in ways that I wouldn’t have predicted. Some of this also comes from natural experimentation: one of my Learning colleagues is using our external DAM as part of a facilitated trail across Brighton Museum with a school group. That’s the last thing I would have expected a DAM to be used for but it suits the needs of this particular session.
From our initial scoping work some immediate actions are underway for our DAM set up.
The internal DAM no longer meets our needs and is unsustainable given the changes in our ICT infrastructure. As such, we are in the process of migrating the assets in our internal DAM to other platforms. Large files, such as TIFFs, are being converted into JPEGs with lossless compression. These have been moved to the external DAM where we have increased capacity. Short-term solutions for the TIFFs and other bulky file formats, such as video files, is being sought pending better knowledge of our future ICT capabilities.
Our external DAM, currently known as the ‘Image Store’ will be renamed and moved to a new URL. The name dates from when it was primarily used for images which were available for sale. As we use a greater variety of file formats and have removed the paywall protecting these images the name is clearly redundant. The URL will shift to something more in keeping with our new website (launched earlier this year) and we are looking at ways to better connect our public DAM offer with our main website.
We are also working with teams across RPM to look at ways of improving the workflows and metadata we use to store our assets.
Big issues ahead
As part of this scoping work we have identified a number of larger issues that need to be addressed. Many of these reflect issues identified in professional guidance such as Spectrum DAM but they are not problems that can be solved simply through the adoption of standards.
RETENTION: We simply cannot afford to keep every digital asset we produce.
Do we really need to keep 40 hi-res photographs of an exhibition launch or learning session forever? How then, do we decide to throw away? It is hard for anyone to consider discarding information that is seen as an investment or an asset, and it’s particularly hard for museum staff and volunteers who instinctively want to protect and preserve the past. I am as sensitive to this as anyone: prior to stepping into my present role I was RPM’s Curator of Photographs. After years of looking after a large collection of photographs it feels odd to be thinking about how to throw stuff away.
SECURITY: Asset Bank contains some good controls for allowing password protected access to specific groups of people for specific assets. However, we need to review if this is sufficiently robust for some sensitive assets such as photographs of children, plans, and other information relating to secure areas of our buildings.
INFORMATION MANAGEMENT: How does our DAM relate to the other information management systems we use? It contains assets from collections, marketing, conservation, learning and many other parts of the organisation; indeed it is one of the few IT systems that is used across the whole organisation. It’s also frequently used as a standalone system and often rivals other systems in its use; our recent survey found that it is frequently used for searching the collections in favour of the data generated from our collection management system.
COMPLIANCE: Some of our assets cannot be made public as they could breach third party copyrights or infringe data protection. While we have controls in place, any system that is geared towards open access and re-use of our assets carries a greater risk if assets are accidentally published. How do we strengthen these controls and procedures without bringing the publication process to a grinding halt?
TRAINING: As we introduce more structured metadata into the system, how reliably can we train staff to stick to these structures? The sheer size of our DAM is even going to make basic searches a complex task. Given that Asset Bank pulls back results against readable search text in documents, a search for ‘Brighton Pier’ is already overwhelming. Staff will need more training in how to set up filters to refine the searches they need.
COST: Even if we develop a retention schedule that strips our preserved assets to an essential core, we will always be collecting more and more data. Even if the cost of storing data continues to fall the production of more digital assets can easily outstrip this. How do we manage and absorb an increasing cost at a time of diminishing resources?
At the moment, I don’t have solutions for everything to hand but potential solutions are emerging from discussions with my colleagues. Our approach will mimic that we’ve used on our new website – both before and after its launch; needs based design, rapid testing, focused discussion with teams around specific problems, and responding to the data we receive about the use of the DAM.
Are we getting this right? We are gradually seeing some positive results from our work so far, but I would be interested in sharing ideas with others wrestling with these same issues.