The last of this morning’s talks (we seem to have fitted a lot in this morning), is by Marshall Breeding, Director for Innovative Technologies and Research at Vanderbilt University. Marshall tracks and has written about the library automation market, some of which is at http://www.librarytechnology.org/
[Wow – there is a lot in this talk, so the notes below are a bit sketchy – I’ll try to pull this together into some more structured thoughts at some future point]
The landscape is that most of the ILS products from commerical vendors are mature – none less than a decade old, and possibly approaching the end of their lifecycle. However, there seems to have been a lack of success in launching new systems (e.g. Horizon 8.0, Taos).
Marshall has put up a slide showing when current systems were architected – the only very recent system is Evergreen (2004) which is open source. The oldest is Unicorn (which we use at Imperial) at 1982.
There has been quite a lot of consolidation in the industry recently (Sirsi & Dynix, Ex Libris & Endeavour) – this is narrowing the choice for libraries (essentially in UK HE there are 4 suppliers – Talis, SirsiDynix, III and Ex Libris). In conjunction with this narrowing of choice, there seems to be increasing dissatisfaction with the products in the market.
The current level of innovation falls below expectations, and companies are struggling to keep up with enhancements and R&D for new innovations. However, some companies are moving forward – not all suppliers are equal.
Marshall is observing that very few libraries change systems except when forced (e.g. supplier goes out of business). It’s hard to justify investments in the ILS to the enterprise/University. There is more money available for next-gen interfaces, federated search, linking, ERM).
A successful pitch for new automation software is one that enables significant transformation toward current visions of the library – you can’t keep doing the same thing in the same way.
We are moving to an age of less integrated systems (this is not just true in library automation) – increasingly we see the ‘core’ ILS supplemented by additional systems (Link resolver, federated search, ERM, etc.)
Many companines involved in library automation are not involved in ILS – e.g. OCLC, Cambridge Information Group/Bowker, WebFeat, Muse Global etc. – none of these produce an ILS.
There is an increased interest in Open Source alternatives – Marshall believes that TCO (total cost of ownership) isn’t significantly different, so perhaps Open Source is a risky alternative – but actual the commercial options also carry risk, so we may just be choosing between different risks.
Some Open Source intiatives:
- Koha Zoom
- Delft Libraries
and in ‘Next Gen’ catalogue interfaces:
So, Open Source is a very small percentage of total picture – but successful implementation breeds confidence and will grow the share. Companies are starting to appear that sell support for Open Source library systems (Index Data, LibLime etc.)
Open Source then is a form of competition for the commercial vendors – which hopefully will lead to pressure to increase innovation, decrease cost, make systems more open, and generally disrupt the Status Quo (in a good way Marshall believes)
To date, the implementation of Open Source in ILS has been based on philosophical reasons – Opern Source will need to compete on a level playing field with realistic ideas of cost etc. to get real traction in the market. Marshall makes the point (which I definitely agree with) that the Open Source systems aren’t actually doing anything different – Evergreen and Koha are modelled on the traditional ILS – it would be good to see more different approaches coming (either from Open Source or Commercial).
We spend ‘at leat half’ of collections budgets on Electronic resources – but the traditional systems don’t help us with this. This is a point I made to Sarah Bartlett yesterday when talking about the Talis ERM project.
So, it seems that libraries are ready for a new approach. Current systems are not fulfilling library need, they are monolithic and complex to administer, and they miss out large areas of functionality (ILL, Book binding, Remote storage management)
Libraries are demanding more openess – this doesn’t necessarily mean Open Source, but open/documented API (beyond proprietary APIs). The ideal is an Industry-standard set of APIs – but this may not be realistic. However there is a current NISO effort to define API for an ILS for decoupled catalogues.
Marshall believes that you can be ‘open’ and ‘commercial’ (I think this will chime with Talis). Looking for Open Data – well documented database schemas, APIs for access to all system functionality. Also more customizability, better integration. Marshall suggests that the key differentiation for vendors will lie in service and support.
A vision of a suite of interoperable modules, with a single point of management for each category of information (we currently do a lot of multiple management – especially with e-journals holdings data) – but not necessarily a single monolithic system. A more lightweight approach – more elegant and efficient, easier to install and administer, automation systems that can be operated with fewer technical staff – the technical team are now dealing with more systems than ever.
The boundaries around the library are getting blurred – online catalogue/library portal/institutional portal – where is the ‘library’. Circulation/ILL/Remote Storage merge, Collection Development/Acquisitions/budget admin; library acq/institutional procurement systems; etc. – all blurring in terms of where boundaries lie.
We are already seeing a clear move to separate the front-end (OPAC) from the back-end with the Next Gen interfaces – this coming from both vendors and open source. This is healthy as (currently) the technology cycle is much faster for the front-end than for the back-end. We don’t want to have to build a new ILS to get a better search interface.
Service Oriented Architecture – this is to some extent what Marshall has been describing – gives flexibility, and this concept is increasing being adopted by the IT industry in general (although there are plenty of sceptics)
Marshall believes we will staft to see massively consolidate implementations – state/province wide ILS implementations, more reliance on consortia, increased use of Software as a Service (SaaS).
Libraries have to both fit within their local enterprise but also the ‘Global Enterprise’ – Google, Google Scholar, Microsoft Live etc.) OCLC Worldcat – why have a local OPAC when you can have Global one? These are issues we need to tackle. Libraries need to leverage the content in enterprise discovery systems to drive users toward library resources – this I think is interesting.
Marshall has given an example of using Search Engine Optimisation techniques (SEO) to get the library records appearing in search engines, driving ‘paid for’ use of the library, funnelling searchers into the library collection.
Marshall is making the point that we are approaching a post-metadata world – the full digital objects are available, and that is what the users are searching – we need to start adapting to this.
We are competing in a crowded field of information providers – commercial web destinations like Amazon have an overlap with services offered by libraries.
So – what can we do? We need to break out of the marketing/consumer model when interacting with commercial partners – we need a substantial dialog that shapes the direction of product development (which is of course what Talis is saying as well). To date we have had 35 years of a evolutionary approach – perhaps we need a revolution – we have to let go of the ILS legacy and find a new model. Web 2.0 has invigorated libraries to think about a lot of the issues – Web 2.0 isn’t the solution, but it has kick-started thinking.