From the analysis of current systems to the conception of future systems : Human Computer Interaction in the French Maritime Patrol Aircrafts
Par Isabelle LEGLISE
ARTIS FACTA - Ingénierie des Facteurs Humains
de IHO-Aéro 1998 Condérence internationale sur l'interaction Homme-Ordinateur en Aéronautique, Montréal, Canada.
Contrary to the traditional methods in the conception of Human Computer Interaction in the domain of civil or military aviation, this paper proposes a process of conception utilizing competencies in Human Factors and in Linguistics. It has been developped in a situation of conception in the navy aeronautics. The process is based on a critical analysis of the existing systems through the analysis of the operators' activity on the systems and of their verbalizations. Improvements for the systems and adaptation to the real needs of the operators have been proposed in order to contribute to the specifications of the future airborne systems, with a special attention to the HCI.
Keywords : conception, HCI, human factors, linguistics, cooperative work, analysis of communication
Introduction : The French Maritime Patrol Aircrafts
The case study we are interested in, takes place in a process of industrial conception. Financed by a French industrial company and the research department of the army, it lasted several years and associated designers, consultants and academics engaged in long-term research projects. The conception of the future airborne systems is currently on hand.
The aircrafts are equipped with various technical systems, such as radars, underwater sonobuoys... enabling them to carry out military missions (missions for information or protection of forces) or civil ones such as rescues at sea. The crew consists of 13 operators: 4 posts in the front (pilot, copilot, 2 mechanics) and 9 posts distributed throughout the plane with various technical specialities.
An unsusual process of conception in the field
Traditionally, the conception of the airborne posts was based on an analysis of the operators' needs, carried out by the designers, thanks to interviews with experts of the field. The latter were "top flight" operators, representatives of the hierarchy or of the competent technical services and "guarantors" of the operational needs.
The study we present is radically different. Our intervention  in the process of conception proposes:
For this study, we attended complete missions, on full-scale simulators (with reproduction of light and sound environment, briefing and debriefing before exercises...) and recorded it (tape recorder 4 tracks and video camera). We also took photographs and collected a certain number of diagrams carried out by the operators during the missions. Lastly, we conducted several interviews with the operators (acoustics experts and tacticians) and also with the instructors.
The analysis of the activity
Our analysis of the activity was carried out with multi-field competences, especially in ergonomics and linguistics. Indeed, the role of the analysis of the communications appeared particularly determining to apprehend the collective activity of the operators  . In the context of a generalizable process , we worked in a very meticulous way on the transcriptions of the communications during the missions and on the recordings of the telephone on board. These transcriptions were, then, processed under the Kronos software, in order to highlight the collective dimension of the activity and to define certain indicators.
On the basis of the analysis of the actions, postures, displacements, gestures or eye positions of the operators, an analysis of their activity was made. Thus, we could model:
The linguistic analysis
The linguistic analysis  attempts to locate, highlight and explain the appearance, in the speech of the operators, of particular traces and linguistic markers such as: a specific lexicon, a verbal time, a mode, markers of subjectivity, a particular intonation, changes of flow, repetitions... The comparison of various verbalization (during the realization of various tasks, and during interviews a posteriori) permitted to find out first tendencies about the linguistic practices in progress on the aircraft: who speaks to whom, about what, how, when...
These elements allowed us, in relation to the analysis of the current systems and to the interviews with the operators, to highlight the limits of the current systems and to make recommendations for the definition of the future Human Machine Interaction. We also took into account the collective dimension of work.
The on board telephone : saturated but essential
A fine study was carried out on the on board telephone (TB). Whereas it is the only prescribed way of communication on the Maritime Patrol aircraft, there are two other types of communications because of the saturation of this channel: shouted communications potentially addressed to all the acoustics operators, and whispered communications, from one operator to the next, over his shoulder. The communications in the TB are audible by the complete staff.
A certain number of failures was noted concerning the TB. Because of its saturation, there are delays in the notification of messages: delayed information does not have the relevance it would have had in real time. Incomprehension, misunderstandings also come from a high information traffic. But lots of questions remain: Which is the role of the TB for the multiple recipients of information ? Is the activity of the latter disturbed by a continuous speech flow in the TB ? Is their situation awareness better thanks to the TB ?
In fact, in order to "clear" the TB, it would be possible to propose automatic sendings of statistical data and automation of certain exchanges by HMI. But all cannot be, in the same way, automated, and certain information must remain and be sent through the TB.
TB: Fine categorizations based on visual observation:
For example, in the exchanges concerning the acoustic data, there are fine categorizations, based on precise visual descriptions, which are, in fact, irreplaceable. These statements, such as noun phrases, are operative   and identify a phenomenon visually observed on the screens of the interface by categorizing it:
apparition de contact, fréquence haute, courte distance
CPA très courte distance sur la 4
fréquence plus basse sur la 8
These categorizations are done on millimeters of signal or on very similar shapes of curves, quite difficult to verbalize.
TB: Collective and certainty
In order to define the type of information conveyed by the on board telephone, a more detailed study was carried out. Astonishingly, the TB miss the pronominal forms of first person, like je. In fact, an operator, seeing an interesting element on its screen announces it, first in a shouted voice with the first person je, then in the TB with the impersonal form on, like:
A3 : -- eh / j'ai XZ sur la 24
A1 : -- ah ouais / ah ouais ouais ouais:
A3 : -- on a XZ sur la 24 (TB)
Actually, on is not only impersonal, it is also the expression of the collective. Even if one operator obtains information on its screen, when information passes through the TB, the information becomes collective.
This passage out of the telephone -> into the TB seems to be related to the certainty of information. Je, subjective by definition, is absent of the TB, just as the forms ouais, which alternate with oui. In a mission, we counted 200 ouais modulating answers to requests for information, and we listed only one of them on the TB, corresponding to a very special moment: one of the operators was requested to come on a decision and he was formulating assumptions. To summarize, the information transmitted by the TB is fine and precise, as certain as possible, and ready to be re-utilized by the collective.
Some recommendations concerning the on board telephone
We proposed the automation of certain exchanges, via data links between the different interfaces, but we took care not to deteriorate the collective work, which requires the construction of common knowledge through communication. Concretely, we proposed the conception of a specific screen dedicated to the presentation of the tactical situation, allowing the acoustics experts to know the position of the buoys released by the tactician and their identification. This source of information must permit to reduce the verbal exchanges and to reach information more quickly and more effectively. In compensation, we alerted the designers about the risk to automate the exchanges concerning, in particular, the visual analysis of the signals detected by the operators.
1. Léglise I., Soulard P. (1997), "Linguistique et analyse de l'activité : une pratique de l'intervention en ergonomie de conception", actes du XXXIIème Congrès de la SELF, 689-703, Sept 97, Lyon
2. Pavard B. (ed), Systèmes coopératifs: de la modélisation à la conception, Toulouse: Octarès
3. Falzon P. (1994) "Functional Dialogues and Collectives Activities", in Le travail humain ndeg.57, 299-312, Paris: PUF
4. Léglise I. (1997), "Intervention linguistique: théorie, pratique et intérêt dans le cadre de l'analyse de l'activité", in revue Linx ndeg.37, 169-182, Université Paris X, Nanterre
5. Faïta D. (1989) "Travailleur collectif et pratiques langagières", in Actes du colloque Travail et pratiques langagières, Paris, ronéo.
6. Hoc J.M., Cacciabue Pc., Hollnagel E (eds), 1995, Expertise and technology: cognition and human computer cooperation. L. Erlbaum Associates: Hillsdale, NJ
haut de page