Title: On The Transformation of Industry Academic Collaborative Research and its Value.
Abstract: Software Engineering Research has transformed the last decade. Not only do most major software producing companies have their own research institutes, in which they have unique access to data and sense of urgency. Problems need to be solved that are important for business, and results does not end up as shelf-ware, they are deployed so fast that the researcher hardly have time to measure the changed outcome. This is good news. We can call it Action Research, but then – the publication trail might not be so important, except for the researcher in academia – often a PhD student – who is left with a difficult task when the industry do not want to spread this “new” advantage to the competition. Yet, more research in industry and more collaboration brings the level of maturity up. All can learn from this. Academic Software Engineering has equally a pressure to produce accurate research that is usable to industry. Software Engineering Research brings value for any engineering when it is applicable. Machine-learning approaches has brought new emphasis is on method, where statistical support to create a basis of research is as it is important in these technique as it is to “know what is right” through training the learning of the set, meaning – to write the right research questions and problems to be solved and have good repeatability. Machine Learning teaches the value of good data. Any software researchers should excel in this approach – since collecting data and asking, “the right” research questions are paramount to valuable research. The issue is “the right” research questions. Does even academia have the right questions for the software engineering research? And how is this basis of knowledge formed? My “field” is software testing. I am astonished all the time how little academics in the field actually know of real software testing (engineering) and how software really is produced. They have never done it themselves. And of course, on that note – I am equally astonished how little the subject is taught at universities – compared to its usage in industry. No wonder there is a “huge” industry for selling test courses based on loose scientific ground. And no wonder the level of test in most industries remain ignorant. There are always exceptions to the rule, but this just shows the need to improve the collaboration between industry and academia? Since – what value should we place on the industry you are collecting data from, do they really know very much? This is the subject of my Keynote.
Short Biography: Dr. Sigrid Eldh is a researcher and senior specialist at Ericsson Radio System & Technology, located in Stockholm Sweden. At Ericsson, she leads the Ericsson research on software test, debugging and product quality. She is an Adjunct Professor at Carleton University in Ottawa, Canada, and a Senior Lecturer at Mälardalens University, Västerås, where she also earned her PhD “On Test Design”. Sigrid has more than 30 years of practical experience from the IT-Industry, with more than 20 years from the Telecom industry specifically. She has started, and chaired some large software testing organizations for practitioners, like SAST, ISTQB, and SSTB. She is now leading the TESTOMAT project, and ITEA3 EU research project on the next level of Test Automation.