Cleanroom experience

Use of the Cleanroom approach has resulted in software with very few errors and does not seem to be any more expensive than conventional development. Cobb and Mills discuss several successful Cleanroom development projects which had a uniformly low failure rate in delivered systems (Cobb and Mills, 1990). The costs of these projects were comparable with other projects which used conventional development techniques.

Rigorous inspection and formal verification seem to be cost-effective in the Cleanroom process. The vast majority of defects are discovered before execution and are not introduced into the developed software. Linger (Linger, 1994) reports that, on average, only 2.3 defects per thousand lines of source code were discovered during testing for Cleanroom projects. Overall development costs are not increased because less effort is required to test and repair the developed software. Selby et al., (Selby, Basili et al., 1987), using students as developers, compared Cleanroom development with conventional techniques. They found that most teams could successfully use the Cleanroom method. The programs produced were of higher quality than those developed using traditional techniques - the source code had more comments and a simpler structure. More of the Cleanroom teams met the development schedule.

Cleanroom development seems to work when practised by skilled and committed engineers. However, use of the method has been confined to technologically advanced organisations and it is not clear how widely it is adopted. While there was considerable interest in this process in the late 1990s, there have not been recent reports of its use.

Reports of the success of the Cleanroom approach in industry have mostly, though not exclusively, come from its developers. We don’t know if this process can be transferred effectively to other types of software development organisation. These organisations usually have fewer, less committed and less skilled engineers. Transferring the Cleanroom approach to these organisations still remains a challenge.


Cobb, R. H. and Mills, H. D. (1990). Engineering Software under Statistical Quality Control. IEEE Software,7(6), 44-54.

Linger, R. C. (1994). Cleanroom Process Model. IEEE Software,11(2), 50-8.

Selby, R. W., Basili, V. R. and Baker, F. T. (1987). Cleanroom Software Development: An Empirical Evaluation. IEEE Trans. on Software Eng.,SE-13(9), 1027-37.


(c) Ian Sommerville 2008