Abstract
Uncrewed Aerial Vehicles (UAVs) within Search and Rescue (SAR) applications can improve the effectiveness and safety of individuals working on the ground. However, the task of detecting a human within the images collected by UAVs is challenging. For this reason, the development of automated image classification aids have the capability to improve the effectiveness of target identification. Such automated technology can be viewed as part of a Decision Support System (DSS) with the potential to improve the decision-making processes of the individuals responsible for analysing UAV imagery in real time. Until now, research looking at the integration of automated functionality has focused on the complexities of piloting and navigating UAVs. As a result, the role of the operator has been somewhat neglected. Given that DSS are yet to be integrated within SAR applications, there is an opportunity to ensure their design meets the user requirements of the operators. This research seeks to understand the decision-making processes employed within a UAV team, comprising a Payload Operator and Pilot, when searching for a missing person.Three types of decision models were investigated: the Recognition Primed Decision Model; the Perceptual Cycle Model, and Decision Ladders. To generate the models, five in-depth interviews were conducted with UAV operators currently working within SAR teams across the UK. The different decision models identified the aspects of decision-making that could be supported with decision aids. From this understanding, each model was used to propose unique design recommendations for a future DSS capable of guiding a Payload Operator's decision-making process. Suggestions for future work are made to ensure DSS development continues to involve the end-user as part of a user-centred design approach.
| Date of Award | 3 Oct 2023 |
|---|---|
| Original language | English |
| Awarding Institution |
|
| Supervisor | Jan Noyes (Supervisor), Seth Bullock (Supervisor) & Victoria Steane (Supervisor) |