On February 1995 LABEIN was awarded the ISO 9001 registration for its software development activities. One of the main comments of the ISO 9001 auditors concerning LABEIN's Quality System was related to the testing procedures, which could be subject to much improvement. The lack of a good testing methodology had been recognised for some time to be a source of high cost in order to obtain commercial levels of quality, and is frequently the cause of budget and/or schedule slips. The project objective is to improve the testing process at LABEIN so that it can be done quicker and cheaper as well as to improve the quality of the final product.
The objective above will be achieved developing the appropriate testing procedures, selecting the most suitable tools that help the developers during the testing phase both in the testing itself and in the testing managing and accounting, refining the method with its application to the pilot project and disseminating the method both internally and externally. The baseline project will be the SRS I and II projects. These are parts I and II of a technical decision-support system running on UNIX workstations which are connected on-line with a SCADA (Supervisory Control And Data Acquisition) system controlling a large electricity transportation network. The SRS-I is a 10 man-year system which has been completed during the first quarter of 1995, while SRS-II is the second part of SRS, a 4 man-year project, and is scheduled to last from July 95 through November 96. Most of the data from project SRS-I are already available, such as effort spent in testing and bugs reported, among others. Since SRS-II includes new functionality plus the deployment of the system at a new customer, real quality data such as bugs found and number of complaints will be available for its use in this PIE, and the improvement in the testing phases of SRS-II due to the PIE will be easily compared with earlier data from SRS-I.
EXPECTED IMPACT AND EXPERIENCE
The main expected benefit of the experiment is the availability of a Methodology which provides support in the accurate dimensioning and management of the testing activities in software development projects, as well as tool support for these activities. Personnel productivity is expected to increase through decreased rework and earlier problem detection. At the same time, an increase of the quality of the products will produce important savings in the maintenance phase, since fewer bugs will be left in the system, which means important savings for the customers and an improved commercial image of the developer. Also, since much of repetitive testing is expected to become automated, the maintenance phase of products is expected to be cheaper and of better quality.