Invited Speakers

Model based testing, and its first cousin, model driven testing have become fashionable fields for research and are generating a great deal of interest in the industry. This survey of the field attempts to define what exactly is meant by the two terms; why the academics and some industrial practitioners are excited by the field; how the tools and methodologies work in practice; and what is needed to create real value behind all the hype.

Our main observation is that the majority of software testing practitioners are focused on the practical value delivered by tools and processes. This means that they don't care if they use model based testing or inter-galactic gobble-de-gook testing. As researchers, we need to focus our energies on providing value to the testers in a palatable form.

We propose a success criterion for industrial adoption of any technology, and look at how model based testing measures up to it.

Dr. Alan Hartman holds a Ph.D. in mathematics from the University of Newcastle, in Australia, and an M.Sc. in mathematics from the Technion Israel Institute of Technology. He joined IBM Haifa Research Laboratory in 1983 after a postdoctoral fellowship at the University of Waterloo in Canada. He has spent two periods on leave at the University of Toronto, and at Telstra Research Laboratories. His career at IBM has included research in Storage Technologies, Communications Network Design, Mathematical Optimisation, and most recently, Hardware and Software Verification. Alan has been involved recently in large scale research projects on model based testing and development including coordination of the EC 5th Framework AGEDIS project, and leading the industrial adoption work in the 6th Framework MODELWARE project. His current responsibilities include management of a research unit at the IBM Haifa Research Laboratory which focuses on the development of new tools for model driven software engineering. Alan has published over 50 research papers and holds 10 patents.

Why is it that some software engineers and computer scientists are able to produce clear, elegant designs and programs, while others cannot? Is it purely a matter of intelligence? What is the problem? One hypothesis is that the answer lies in abstraction: the ability to perform abstract thinking and to exhibit abstraction skills. Abstraction is a cognitive means by which engineers, mathematicians and others deal with complexity. It covers both aspects of removing detail as well as the identification of generalisations or common features, and has been recognized as a crucial skill for software engineering professionals. Is it possible to improve the skills and abilities of those less able through further education and training? In this talk, we explore these questions, discuss why abstraction tends to be ignored and argue that modeling provides a sound means for using and improving abstraction skills.

(Wikipedia: The elephant in the room is an English idiom for a question or problem that very obviously stands, but which is ignored for the convenience of one or other party. It derives its symbolic meaning from the fact that an elephant would indeed be conspicuous and remarkable in a small room; thus the idiom also implies a value judgment that the issue should be discussed openly. )

Professor Jeff Kramer is Head of Distributed Software Engineering in the Department of Computing at Imperial College. He was Head of Department from 1999 to 2004. He was a principal investigator in the various research projects which led to the development of the CONIC environment for configuration programming and the Darwin architectural description language which is used in commercialised form by Philips for the software for high end television sets. His current research work is on behaviour analysis,the use of models in requirements elaboration and architectural approaches to self-organising software systems.

Jeff Kramer is a Chartered Engineer, Fellow of the IEE and Fellow of the ACM. He was program co-chair of the 21st ICSE (International Conference on Software Engineering) in Los Angeles in 1999, Chair of the Steering Committee for ICSE from 2000 to 2002, associate editor and member of the editorial board of ACM TOSEM from 1995 to 2001 and is currently editor-in-Chief of the editorial board of IEEE TSE. He was awarded the IEE Informatics Premium prize for 1998/99 for a paper on Software Architecture, was winner of the Most Influential Paper Award at ICSE 2003, and was awarded the 2005 ACM SIGSOFT Outstanding Research Award for significant and lasting research contributions to Software Engineering. He is co-author of a recent book on Concurrency, co-author of a previous book on Distributed Systems and Computer Networks, and the author of over 150 journal and conference publications.

Despite decades of academic and industrial research in testing and analysis of software, few tools and ideas have found their way into the programmers' toolset. Many find this to be a concern. Are the researchers asking the wrong questions? Is their work not relevant to real-world problems? Or, are they simply not successful in promoting and commercializing their technology? We do not think that any of these conclusions are entirely true.

Much of the research in testing and analysis has focused on important and fundamental scientific questions, providing limited opportunities for direct technology transfer. The time is ripe, however, to apply the results of this research to create more practical methods and tools. We believe that creating successful methods and tools that can serve as the basis for technology transfer requires the researchers to focus on three fundamental issues:

• Finding problems faced by the majority of industrial software development and finding solutions that developers will embrace.
• Knowing when a research result is ready for transfer and knowing what constitutes a transferable asset (a tool? an idea? a body of knowledge?).
• Choosing and applying an appropriate model for transferring the research results to an organization that will undertake further development (and possible commercialization) of the technology.

This talk will describe these three constituents of successful technology transfer, present some success and failure stories, and summarize some of our observations from over a decade's worth of experience in deploying testing and analysis tools "in the trenches."

Marat Boshernitsan is the Director of Research at Agitar Software, Inc., where he pursues his passion for building software development tools that help programmers create higher quality, more maintainable software. In the past Marat has worked at SunLabs, where together with Michael Van De Vanter he developed CodeProcessor, a program editor that provided typographically-grounded automatic formatting to improve readability of source code. Marat received his Ph.D. in Computer Science from the University of California, Berkeley under the direction of Susan Graham. As part of his dissertation research he developed a system for automating systematic source code editing through interactive creation and execution of formally-specified source code transformations.