Authors:
Turkan Hentati
1
;
Théodore Letouze
2
;
Charles Dormoy
1
;
Jaime Diaz-Pineda
3
;
Ricardo Reis
4
;
Anaisa Villani
4
and
Jean-Marc Andre
2
Affiliations:
1
CATIE, Bordeaux, France
;
2
Bordeaux INP-ENSC, IMS UMR 5218, Université de Bordeaux, CNRS Talence, Bordeaux, France
;
3
Thales AVS, Bordeaux, France
;
4
Embraer Research and Technology Europe - Airholding S.A., Alverca do Ribatejo, Portugal
Keyword(s):
HAT, Bidirectional Communication, Decision Making, Intelligent Assistant, Human-Cooperative Techniques.
Abstract:
Commercial aviation is increasingly constrained by airspace congestion and the need to balance profitability with environmental concerns. Despite this growing complexity, pilot’s cognitive resources remain the same. This article examines a new communication paradigm using 'intentions' in HAT (Human Autonomy Teaming) for commercial aviation. The use case involves a cockpit IA (Intelligent Assistant) designed to assist flight crew in re-routing or diverting an airliner to a new destination in the event of weather hazards, taking into account various operational performance indicators. To communicate and negotiate with the IA, the pilot expresses their high-level goal, also known as operator intention, which includes preserving cognitive capacities, passenger comfort, or airline profitability, in order to find the optimal solution. This work compares three types of assistance: decision support, cooperative assistance, and collaborative assistance. The study aims to identify the key feat
ures of each type and determine the most suitable level of assistance for supporting decision-making during rerouting. To validate the objectives of this use case, six pilots were asked to evaluate three different types of assistance using the 'cognitive walkthrough' method and questionnaires about trust and usability. The results provide some key features of each type of assistance that can increase the performance of decision making in a distributed work between pilot and IA.
(More)