The E3million project ran for more than two years. The organizations that partnered for the project included IBM, the University of Southampton, MTA SZTAKI, and the German Aerospace Centre. With potential applications investigated ranging from the aerospace industry to organ donor management, the project reveals the potential universal appeal of the application of provenance.
Legislation has already required several industries (such as financial services and agriculture) to create one-off solutions to the requirements for traceability. By the nature of such proprietary solutions, implementation is extremely costly, and ongoing maintenance more so. It would be reasonable to expect legislation to impose further traceability requirements, increasing both the depth of reporting and the breadth of applicability. The need for a standardized approach and commercial products is evident.
The approach taken by the Provenance Project has been to start by addressing the need for a common data model that will give rise to generalized search, analysis, and visualization products. As might be expected, the data model is defined as a set of XML schemas, and described in a set of specification documents. Although these specifications have not yet been presented to a standards group, they are open for inspection and download at www.gridprovenance.org.
The project created a data store for provenance data using the open source eXist XML database. A client-side library was also created that provides access to the data store via an API. Several follow-on activities are envisaged including:
* Prove the interoperability of provenance data, initially using import/export, but with the future inclusion of remote access and security constraints.
* Build a set of potential user sites that would like to take part in trialing proof of concept products.
* Create toolsets to simplify the collection of provenance data from application services.
Content from our partners
Because the activities of interest will typically take the form of message pairs that invoke a service and communicate the results, provenance technology is likely to prove to be a very good fit with service-oriented architecture (SOA) and its use in automating business processes. Ultimately, the collection of provenance data could be implemented as a customizable plug-in to an enterprise service bus.
There is plainly a growing need for traceability of information and of activities leading up to a decision or event. It is a particularly challenging task to generalize this to the extent that commercial products can replace bespoke development of proprietary solutions. It is a little unfortunate that an academic language is being used that will require interpretation before the business benefits become obvious.
Recent surveys of delegates at Butler Group conferences have confirmed that it is high time that IT and business agreed on a common language so that meaningful discussions can take place on the appropriate use of technology. While this Provenance Project is plainly still at an early stage of evolution, it would be much easier to obtain the business commitment needed for the next phase had a less academic language been used. Despite this, the growing demands of legislation make it likely that this project will ultimately result in standards-based commercial products that will simplify, and lower the buy-in price, for a growing number of traceability projects.
Source: OpinionWire by Butler Group (www.butlergroup.com)
