| Manny Lehman |
The Software Process Group focuses on the following topics in Software Engineering: Software Process, Software Evolution, Feedback in the Process, Process Improvement.
A Simplified View of a Software Evolution Process. From "Model-Based Assessment of Software Evolution Processes" by G Kahen, MM Lehman and JF Ramil.
The research addresses the above topics seeking to model the global software process as a feedback system and analyse its behaviour. The goal is to develop theories of software process and software evolution, producing models of such processes and the evolution phenomenon, applying the models to support software process improvement, providing tools for software release planning, management and so on. The global process includes both technical aspects and non-technical aspects such as, for example, project management, organisational management, marketing, user support and user activities like submission of comments, fault reports and change proposals. Some can be formalised, others not. Following the successful conclusion of an EPSRC supported, two year project FEAST/1 which produced, inter alia some 25 publications (not including research reports), the group is now half way through FEAST/2 having made further significant progress. A paper discussing the immediate practical application of the results is in the final stages of preparation. A proposal to develop a formal theory of software evolution is being prepared.
The investigation began by studying, modelling (black box and white box) and interpreting the evolutionary behaviour of appropriate systems developed and supported by the industrial collaborators (ICL High Performance Systems, Logica, Matra-BAe, Lucent Technologies, MoD DERA). This led to significant support for a majority of the results and observations obtained by the principle investigator in earlier studies in the seventies, in particular most of the Laws of Software Evolution and the Software Uncertainty Principle. The confirmation of the presence and significant impact of feedback and systems dynamics effects is facilitating the construction of system dynamics models with a view to identifying and learning to control the global process feedback mechanisms. Further support for the Laws and the principle and the insights that underlie them provide the basis for the theory mentioned above. A workshop with international participation, FEAST 2000, to discuss past results and review plans for future research and development in this area is to be held in the Department in July 2000.
The focus on the global process, unique in the field of software engineering process research, implies that the results of this work may be expected to have wide potential application in business and organisational process improvement and in the socio-managerial aspects of computer application.