Umu logo Umeå universitet
Teknisk-naturvetenskaplig fakultet
Institutionen för datavetenskap

PROJECTS


Quality Management for Small Enterprises

Jürgen Börstler
Department of Computing Science
Umeå University, Sweden

Olof Johansson
Svenska Test
Umeå, Sweden

Terttu Orci
Department of Computer and Systems Sciences
Stockholm University/KTH, Sweden

Background

It is evidently difficult to develop software systems in time, within budget, and with the agreed quality. During the last two decades, the productivity of software engineers has increased with orders of magnitude, the educational level of them and of the managers has certainly improved, new methods, techniques, languages, and environments have been introduced with a rapid pace. Surprising or not, the same characteristics still apply to software engineering projects: they are often late, over budget, and the product does not satisfy the customers' needs.

The software engineering community seems today to commonly agree that the focus should be shifted from single methods, tools and languages to efforts towards improving the entire development process including project management. The process improvement models like CMM [PAU95] from USA, BOOTSTRAP [BOO97], SPICE [ROU95] at the European level, and standards, e.g ISO9000 [OSK95], they all are results from this awareness.

While the above models are intended to be used at an organisationwide level to improve software processes, there are also models focusing on individuals and teams, namely PSP (Personal Software Process) [HUM95], [HUM97], and TSP (Team Software Process) [to appear in 1998], and also a model for improving human resource management P-CMM, the People CMM [SEI]. PSP is a model for improving the work of the individual software engineers: the development, measurement, and estimation. TSP is the corresponding approach on the team level. P-CMM is a model for human resource departments in organisations developing software, although the generality would probably generate the benefits also for other type of organisations.

It is important that the individuals, teams, and management, all agree on an improvement to be undertaken, otherwise the improvement programme will face difficulties to survive.

The Problem

The existing models for process and people improvement, have proved to deliver their promises in large and medium size companies. However, to get the education and training in these models, and further to get the certificate, is costly both in nominal terms, but also in terms of the effort required, and lost income during the training and introduction time. Certainly, the large companies can take the costs and time needed to systematically implement all the actions needed to gradually climb on the improvement ladder, whether it will be CMM, P-CMM, PSP for TSP. In small enterprises, however, with 5-15 employees, the situation is different: it may be difficult to motivate the time and costs, the manager may be a member of the development team, thereby having a double role. The organisation structures, proposed in CMM and also in other improvement programs, are impossible to implement without modification in small companies. For example, senior management, QA, SEPG, PAT, are roles requiring a possession by a person. One and the same person might need to accept the role of SEPG member, QA, and a PAT. Naturally, the basic idea of the roles will then diminish.

A company starting with a couple of software developers may function very well: all the developers have oversight and natural communication. If the company is successfull, it will usually grow, with new employees hired. At some point in time, the oversight and natural communication will be lost. The company, especially if successful, will have increasing number of contracts, with the consequence of increasing profit, increasing staff, and also with the existing staff working over time to rescue projects that are late. Usually, it is well understood that something must be done in terms of people management and process improvement, but in a chaotic process, it is difficult to take the time to do the improvements, just to get the current job done.

Especially, for the small companies, the improvement efforts should become a way of life and organisation culture before the oversight is lost and there is a chaos. To this end, quality management support is needed for small companies, before they grow.

Related Work

The importance of process improvement efforts on the European level is acknowledged by the introduction of the EC program ESSI (European Stratetic Software Initiative) the goal of which is to support projects with software improvement efforts in organisations, preferably small and medium size companies. Companies can apply for financial support (50%) of the costs of undertaking a process improvement experiment (PIE). Until today, more than 300 PIEs have received support from ESSI programme. None of the more than 300 PIEs have concerned implementation of CMM, rather they all start from some of the lifecycle areas, e.g. configuration management, design, object-oriented technology for analysis and design.

All the PIEs can be seen as related work to this project: they implement an improvement effort, including measurement and action. The lessons learnt by the PIEs can be utilized in this project. There are/have been several projects, so called ESSI dissemination actions, trying to collect the lessons learnt and to disseminate the experiences of the individual PIEs. One of those is EUREX (European Experience Exchange), where of the applicants (Terttu Orci) is participating. The idea of collecting and disseminating the experiences is a good one, in practice the work is, however, very hard, as it is based on second hand experience, on the final reports of the PIEs. The quality of the EUREX reports can never be better than the PIE report quality. Therefore, we doubt that individual small and medium size companies can directly make use of the PIE experiences, nor undirectly by looking at the dissemination actions. They give, at best, some useful hints, but more useful general guidelines and support are needed.

Expected Results

The long-range goal of this project is to develop support for quality management for small enterprises. The intention is to tailor the existing models to small companies, and to propose a model with the integrated approach, i.e. the work in this project will proceed on breath and on depth. The expected results in long term will be a practically and theoretically sound quality management model for small companies, with adequate tool support.

Work Breakdown Structure

The work breakdown structure will be defined for the first project year 1999. At each task, there is a milestone defined. Each of the milestones will be made available for the project control board for review. The work in the following years of the project will continue with the same approach along the higher levels, 3-5 of the source models.

References

[BOO97] http://www.etnoteam.it/bootstrap/institut.html
[EUR97] EUREX - European Experience Exchange, Project No 24478, 1997.
[HUM95] Humphrey Watts S: A Discipline for Software Engineering. Addison Wesley. 1995.
[HUM97] Humphrey Watts S: Introduction to the Personal Software Process. Addison Wesley. 1997.
[OSK95] Oskarsson Ö, Glass R: ISO 9000 i programutveckling. Studentlitteratur. 1995.
[PAU95] Paulk Mark C et al: The Capability Maturity Model: Guidelines for Improving the Software Process. Carnegie Mellon University, Software Engineering Institute. Addison Wesley. 1995.
[SIS96] ESPITIs enkätundersökning i Sverige. Svenska Institutet för Systemutveckling SISU, 1996.

http://www.cs.umu.se/~jubo/Projects/QMSE/application.html
Last modified: Wed Jan 13 11:07:49 MET 1999 by jubo.