Develop a Highly Usable VR System - and Do It Yesterday!"

The project had identified a number of key areas which needed specialist attention. These were to:

• Produce a definitive statement of requirements on the DISCOVER system from the training organization. This involved understanding the methods used in current training, which ranged from state-of-the-art physical simulators costing tens of millions of pounds, to simple role playing, and their methods of assessment.
• Adopt a suitable pedagogic model for the training environment.
• Understand how we were to validate the DISCOVER system with the appropriate validating bodies. The project had ambitions to sell the system across the world to a number of shipping and oil companies and thus required a suitable seal of approval.
• Produce for the software designers a detailed description of the virtual environments to be modelled and a list of the objects and their behaviours therein.
• Specify a collaborative virtual environment with a strong emphasis on fidelity and usability.

Match the system to current and proposed training scenarios. And we had three months to complete this in the first instance!

Phase 1: Just What Should a VirtualTraining System Do?

Some requirements work had been carried out before our entry to the project, but this had simply taken the form of the collection of high level statements of intent from the training companies and lists of competencies (or skills) to be trained. This provided our first political advantage: working closely with managers and training personnel in the user companies in itself convinced these project partners that their needs were being seriously considered. Thus our activities tended to be viewed in a constructive and receptive way. We will not provide a detailed description of the techniques we used, since they are familiar ones, but we employed a combination of interviews with trainees, trainers, their managers, observation and video- recording of training sessions and the collection of relevant documentation. All this took place at user sites. Again, goodwill was generated simply by the amount of effort very clearly being devoted to the user-centred work.

The results of this fieldwork fed into a large group of requirements relating to the way in which CVE based training could best be used in its particular intended context. Fortunately, although CVE design is a young field, there isan emerging body of literature which addresses generic usability issues for single users (e.g. Stanney et al., 1998; Sutcliffe et al., 2000) and in the collaborative context, the guidelines provided by the COVEN project (Normand et al., 1999). Thus we were able to supplement our own primary findings with secondary material from the literature and produce a large body of user-related requirements material in a relatively short time.

The Use of the MoSCoW Method

At this stage we had a large set of requirements in clear need of prioritization. In keeping with the user-centred philosophy which we had started to foster, this would be done in conjunction with the user companies. The MoSCoW technique from the Dynamic Systems Development Method - DSDM, (Stapleton, 1997) afforded a means of doing this which would be readily understandable to users and familiar to the developer partners in the project. DSDM prioritizes requirements using the MoSCoW rules.

The uppercase letters of the acronym (the o's are only there to make for a memorable acronym) stand for: Must have for requirements that are fundamental o the system. Without them the system will be unworkable and useless. The Must have category defines the minimum usable subset. Should have for important requirements for which there is a work-around in the short term but the system will be useful and usable without them. Could have for requirements that can more easily be left out of the current development. Want to have but won't have this time round for those valuable requirements that can wait till later development takes place. A small sample of the prioritised requirements follows below A substantial subset of the higher priority lists related to usability and other non-functional concerns.

• lMust: The ability for trainers to increase stress in a safe, controlled manner for the trainees.
• Should: Provide a more satisfying professional role for trainers than classroom teaching.
• Could:Courses formally validated by the appropriate standards bodies.
• Want：Measures of trainingtransfer.

By now, therefore, we had an substantial body of organized and prioritized requirements which formed the main body of the requirements and evaluation deliverables to the funding body. These reports passed their review, user-related concerns were accepted as the main force driving the design of the software and the project moved forward.