Article Preview
Top1. Introduction
The study of controllers’ work performances within their natural work environment is an activity rewarded with innovative requirements which cannot be easily achieved from simulated platforms and controlled laboratory settings. However, a major hurdle in productive requirements elicitation is the a-priori lack of specialist knowledge concerning the operational environment, opposed to the necessity of ‘asking the right’ questions as soon as possible. Although it can be argued that an experienced practitioner in Air Traffic Management (ATM) should minimise this delay, variations in control room management strategies, work techniques and other matters can easily disrupt the productivity of the requirements elicitation process.
Requirements elicitation aids are available to a practitioner. However, they have the burden of providing sufficient adaptability as a means of addressing the problem of unexpected changes in the target operational environment and during the period of a field study. We can illustrate the problem by imagining a practitioner utilising basic note-taking methods for collecting in-situ data. In the more conservative case, the practitioner enters the field using blank pages, which are then augmented using textual descriptions and graphical depictions of the work environment. However, the method is time consuming and might incur certain redundant information pertaining to stable features of the work environment. In the alternate extreme that the practitioner enters the field using a more prescriptive protocol, the problem is that of recording novel work practices which have not been structured a-priori, by the protocol.
Thus, a finer spectrum of recording methods is required for eliciting requirements in dynamic work environments. This paper reports on the parallel development of a physical aid to requirements elicitation, as well as the theoretical design rationale constructed within the context of field studies in ATM. We pose the following research question: what are the factors influencing the design of aids during requirements elicitation and how is the design process rationalised?
The specificity of this research question needs to be contrasted with similar initiatives. First, we are interested in the design as a purely emergent activity, as compared to a specified process. Thus, design as a consequence of a knowledge-sharing goal is of interest, rather than a central goal in itself. Second, the functions of design and usage are shared between the same actors, rather than dedicated. According to this treatment, the design activity is distributed. Third, the classical process of requirements elicitation, designing and validation, is tightly coupled in time, ranging from a cycle of a few minutes to a few days. This is contrasted with lengthier design methodologies engaging a cycle of weeks to months. While several concepts come close to those initial considerations, they do not satisfy all of them.
We make use of Computer Supported Cooperative Work (CSCW) framework as a means of characterising the development of representational forms between practitioners and operators(Schmidt & Simone, 1996). In CSCW, work is construed as a collective, social activity supported by technological means; workers are called actors given their integrated participation towards goals; specialised groups of actors whether specified or emergent are called communities of practice (CoP) (Wenger, 1998); interaction between such CoP are mediated by special artifacts called boundary-spanning or boundary-negotiating objects (Lee et al., 2005). The cooperation within and between CoP involve different information exchange processes, among which are ordered management of information exchanges, called coordination (Malone, 1988).