Software Engineering Institute goes International in Software Process Research
by Mario Fusani
In industrial prototyping and production, the shifting of the object of study from product to process is more than a century old and has been at the basis of significant product improvement. In software technology, a similar effort has been attempted for a couple of decades, both with empirical software engineering approaches and with mathematical models, yet the results in software-dependent products and services are not as bright as could have been expected.
On the empirical side, corporate and public standards for software development have flourished in recent years. It has been shown that standards impact directly on industry (even the software industry) much more than solutions proposed in specialised literature. For this very reason, standards should embody applicable results of scientific research, whereas frequently they do not. A consequence is that, despite an increasing tendency of developers to follow recognised international norms, there is growing concern about the ability of empirical software development processes to respond to the emerging needs of the global information society. On the formal methods side, indicative is the fact that the application of mathematical models (together with the instruments that can be derived) by software process managers is still a somewhat rare occurrence. So the big questions are:
- what will be the successful paradigm(s) for the software process in the near future?
- what requirements must the software satisfy?
- how will these requirements be influenced by other (social, political, economical, geographical) factors?
At the Software Engineering Institute (SEI), this concern has been taken on-board as an exciting challenge (the bigger the problem, the happier the researcher, one could say). 'Exploring What's Next in Software Process Research' has been the leading concept of a recent initiative, sponsored by SEI and by several private corporations and has led to the constitution, in August 2004, of the International Process Research Consortium (IPRC), a worldwide association of researchers. Six European countries are joining SEI in this effort, Italy being represented by a researcher from ISTI-CNR.
The main goal of IPRC is to prepare a Roadmap for software process research for the next 5-10 years. The aim will be to provide indications on how the technology challenges of the near future should be addressed by the software industry. However, software is not the only issue of interest for IPRC researchers. Here below we summarise the main activities:
- a number of working groups have been set-up to investigate distinct areas of interest
- six workshops will be held between August 2004 to August 2006 in which ideas of the IPRC member researchers will be proposed and debated; working groups will continue their activity in the meantime
- different nationalities, cultures and needs are represented in the research teams, to avoid localisation and privileged solutions
- experts in non-technical disciplines, such as economy, psychology and organizational science, are invited to lecture to the technical team
- experts (from SEI) in working group behaviour act as facilitators, catalysing reactions among researchers
- various interaction mechanisms are activated by the facilitators to obtain maximum benefit from the collaboration;
and the current IPRC approaches and contents:
- technical and non-technical discussion has been solicited from the members in an unbiased way, allowing considerable divergence of points of view and perspectives; convergence is expected as ideas and activity mature around elements of the roadmap
- widely-disparate content has been deployed so far in the form of statements in order to establish relationships and priorities: needs, hot/warm/cold research topics, state-of-art and trends in software process solutions, human factors, known and mysterious forces pushing forward/ retarding progress, and so on
- a scenario-based approach is being explored to analyse the impact of a future multi-dimensional 'trends space' in any of the software process elements; the lack-of-crystal-ball uncertainty is amended by considering each trend in two opposing directions; this implies much work if completeness is the goal, but knowledge can be increased, even with partial analyses.
The outcomes of the two workshops held so far are considered very promising, thanks to the huge amount of material examined and the ideas generated by the discussions. Although much work is needed before there can be convergence towards concrete proposals, we already have the impression that, in addition to the Roadmap, other results and new lines of research will emerge from the work of the Consortium.
Mario Fusani, ISTI-CNR, Italy
Tel: +39 050 315 2916