REAIMS (Esprit Project 8649)

Requirements Engineering adaptation and improvement for safety and dependability

Objective

To develop a framework for improving the requirements engineering process for dependable and safety-related systems and to assess this in a real industrial context.

REAIMS was a 2-year project, partially funded by the European Commission. It started on 1st March 1994.

REAIMS Publications

A number of project publications are available in both PostScript and Acrobat format.

Details are also available regarding the book: Requirements Engineering, a Good Practice Guide, which was produced as a result of the REAIMS project.

The REAIMS Family

In addressing the four strands of activity in REAIMS outlined below, project partners have developed a number of modules and good practice guides addressing the various techniques covered in the overall REAIMS process improvement strategy. Further information on these modules is provided in a more detailed overview of the project. Alternatively, you can access the descriptions and contacts for each module from the links below.

REAIMS Improvement Model		REAIMS Focus
RE Good Practice Guide		PI Good Practice Guide
PREview-RE	PREview-PV	PERE	MERE	FRERE

Workplan

REAIMS was an application-driven project whose main aim was to allow organizations to develop and assess requirements engineering processes for dependable systems. It proposed mechanisms for process improvement by suggesting methods for structuring the requirements elicitation and analysis process, used social studies of errors and accidents to provide guidelines for the design of dependable processes, investigated the introduction of formal methods in the specification process, and developed a generic maturity model for requirements engineering. All of the research was informed by industrial needs and was evaluated in real applications in the railway signalling and aerospace domains.

Important characteristics of the REAIMS project are that it was not just concerned with software engineering but concerned with systems requirements engineering. It also focused on safety-related systems so was concerned with developing processes which recognise the importance of safety concerns and which are designed to include appropriate safety and dependability analysis.

Activity in the REAIMS project was grouped into four distinct strands:

1. Process improvement objectives and techniques

This work was concerned with defining process improvement objectives and methods of achieving these objectives. This involved examining existing processes, developing process models and analysing the advantages and disadvantages of these existing processes. Psychologists and social scientists were also involved in studying human errors and their causes and the influence of social factors on dependable processes. We investigated the structuring of the requirements and the requirements elicitation process using a concept known as viewpoints (an extension of the original viewpoint concept developed in the CORE method).

2. Requirements process improvement through the reuse of expertise

This work was concerned with understanding and improving the requirements engineering process at Aerospatiale. This involved developing methods and procedures for capturing and reusing experience which was previously held informally, with modelling the existing requirements engineering process and with improving the existing RE process by, e.g., incorporating the notion of viewpoints. The revisions were evaluated in the context of a real aircraft programme.

3. Integrating formal methods with the RE process

This strand was principally concerned with investigating the use of formal methods (the B method) in the specification of railway signalling systems and with integrating the work on viewpoints and human factors analysis with a requirements engineering process which includes the use of formal system specification. Again, the process improvements were tested on a real application from the railway signalling domain.

4. Requirements process improvement and maturity modelling

This work was concerned with generalising the research on viewpoints and human factors and adapting it using feedback from the evaluations in the industrial applications. We proposed guidelines for process improvement and a requirements engineering process maturity model. This will allow organizations to assess their process and will serve as a framework for process improvement. Given that the work is generic rather than specific to a particular organization, the model does not propose any specific process but identifies desirable process attributes.

An important objective of the REAIMS project was the exploitation of the work outside of the partners involved in the project.

REAIMS People at Lancaster

Ian Sommerville
Pete Sawyer
Stephen Viller

For further information, please contact:

Ms Chris Needham,
Computing Dept.,
Lancaster University,
LANCASTER LA1 4YR,
UK

Email: chris@comp.lancs.ac.uk
Phone: +44-(0)1524-593041
Fax: +44-(0)1524-593608

Page last updated Thu, 18 December 1997. Please report problems to reaims-request@comp.lancs.ac.uk