This http://equipment.data.ac.uk/ service auto-discovers open data about university equipment in the UK and provides a search interface and also a combined dataset for reuse. The mechanism of auto-discovery we have developed for this can be used for many other similar applications and is a decentralized solution which means that even if the central service breaks or goes away, other people getting value from the data can still obtain it directly from the source organizations. All of the tools we have developed for this are open source and available at https://github.com/data-ac-uk/ . The service promotes interoperable and open data principles and has over 40 UK institutions contributing data. It is fully automated so very low cost to run day to day. The resulting service helps UK researchers collaborate and discover existing equipment to save taxpayer's money being wasted.
Once a week all *.ac.uk homepages are scanned for auto-discoverable open data profiles. Those discovered, with equipment datasets listed, are added to a list. This list of datasets are downloaded once a day and combined into a joint dataset which is published and also may be searched. Errors in the datasets are handled per-record so that an organisation does not disappear due to one mistake. Issues are listed at http://equipment.data.ac.uk/status and also provided in a machine readable format if a member wants to create automated monitoring/alerting of any issues.
The solution could easily scale to massively larger as it's so straight forward. The profile document can be used for any data which is useful to aggregate from many distributed sources. It could be useful for museums, other university applications, doctors and hospital wards.
The service is based on the following technologies: PHP, Linked Open Data, OPD auto-discovery mechanism (our own innovation, details at httpd://opd.data.ac.uk/)
We use a meta-data schema called uni-quip which was developed as part of an earlier projects. We have registered the auto-discovery method with IANA http://www.iana.org/assignments/well-known-uris/well-known-uris.xhtml (openorg). We use RDF/Linked data and URIs to identify everything. We cross link the information with another of our services http://www.iana.org/assignments/well-known-uris/well-known-uris.xhtml.
This project was developed by the University of Southampton, but we are handing over the day to day running of the service to Jisc https://www.jisc.ac.uk/ as they are well placed to look after it. We are continuing to explore other ways the approach we developed could be used.
Awareness of what equipment UK universities have which can aid strategic planning and avoid costly duplication.
It allows researchers and industry to find who has equipment for collaboration or direct contracted work. It also allows people considering buying expensive equipment to find peers who already have it and find out if it's as good as the supplier says. By getting universities to come up with these lists we also help staff find out what their own university already has. Most staff don't know what's in buildings outside their department and before our project they had no way to find out.
The basic system is so automated that new universities can add themselves with no intervention. The mechanism of getting data from institution's websites means that we don't need to maintain a record of authorised contacts at each organisation which would have made the service unsustainable.
We still expect to make the odd improvement via github.
More information is available in a paper written by our project manager: http://www.mdpi.com/2220-9964/5/10/167/htm