Embedded Software: Industry versus Research

This paper summarizes the results from a Danish study of d e v e l o p m e n t methods used for embedded software. It compares the needs of industry with the apparent lack of influence from Danish research. The main conclusions from the study are: (1) Industry have --many problems related to development of embedded software, for instance ensuring that essential issues are dealt with early and not detected at the end of testing. (2) Only very few research results are transferred to industry. The lack of technolQgy transfer is primarily caused by the researchers' lack of appreciation of the real problems in industry. (3) Even very good and very relevant results usually fail to be taken into regular use in industry. Many different barriers cause this failure. Finally the paper suggests h o w t h e s e technology transfer problems could be overcome.