Keywords: Cognitive System Engineering, Human Machine Systems, Supervisory Control
Abstract: The Air Traffic Management (ATM) system can be defined as a “Joint Cognitive System” of people, teams, and artifacts that adapts to the challenges and demands posed by familiar and unfamiliar situations in a dynamically evolving operational context. In the era of digitalization and Big Data we live, an incremental modernization of the ATM system is expected in the coming years with the pervasive implementation of Artificial Intelligence (AI) and Machine Learning (ML). In this paper, we present the findings from an initial attempt to detect and document the fundamental challenges of the introduction of AI, in the European ATM system through the lens of Cognitive Systems Engineering paradigm. We also discuss how these challenges give rise to difficult to resolve safety and performance related patterns in the ATM system.

Keywords: Human Machine Systems, Human – Computer Interaction, Cognitive System Engineering
Abstract: Air traffic control (ATC) is a complex and demanding job reserved for highly-trained professionals. Training ATC candidates is challenging as trainees are subjectively assessed by instructors who are biased by their own ways of working. As an effort to determine control expertise objectively, this study employed clustering techniques on an existing data set in which course and professional controllers participated in a medium-fidelity simulation experiment. Results identified a set of eight measures that formed two distinct and stable expertise clusters. A subsequent sensitivity analysis was able to reveal how far (or close) each course participant was positioned from the expert cluster and on which measures those participants deviated from the experts. At this stage, however, it is difficult to translate these results into specific advice on how to improve underdeveloped skills. Despite the small sample size and limited generalizability of the results in this exploratory study, the method appears to be a promising demonstration in determining objective factors that describe ATC expertise, warranting further research.

Keywords: Supervisory Control, Human – Computer Interaction, Human Machine Systems
Abstract: To meet increasing safety and performance demands in air traffic control (ATC), more advanced automated systems will be introduced to assist human air traffic controllers. Some even foresee complete automation, with the human as a supervisor only to step-in when automation fails. Literature and empirical evidence suggest that supervising highly-automated systems can cause severe vigilance and complacency problems, out-of-the-loop situation awareness and transient workload peaks. These impair the ability for humans to successfully take over control. In this study, situation awareness prompts were used as a way to keep controllers cognitively engaged during their supervision of a fully automated ATC system. Results from an exploratory human-in-the-loop experiment, in which eight participants were instructed to monitor a fully automated ATC system in a simplified ATC context, show a significant decrease in workload peaks following an automation failure after being exposed to high-level SA questions. Although the selected method did not necessarily yield improved safety and manual control efficiency, results suggest that using situation awareness feedback in line with controllers’ attention could be an avenue worth exploring further as a training tool.

Keywords: Human-Machine Interfaces, Design, Analysis, and Evaluation, Machine Learning
Abstract: Studies have shown the possibility to classify user tasks from eye-movement data. We present a new way to determine the optimal model for different visual cognitive tasks using data that includes two types of visual search tasks, a visual exploration task, a blank screen task, and a task where a user needs to fixate at the center of any scene. We used CNN and SVM models on RGB images generated from fixation scan paths from these tasks. We also used AdaBoost on filtered eye movement data as a baseline. Our study shows that deep learning gives the best accuracy for classifying between visual search tasks but misclassified between visual search and visual exploration tasks. Machine learning-based methods performed with high accuracy classifying tasks that involve minimal visual search. Our study gives insight on the best model to choose by type of visual task using eye movement data.

Keywords: Model-Based Design, Human Machine Systems, Human – Computer Interaction
Abstract: Design proposal for an AI-driven military situational awareness application. Current events in Ukraine have emphasized the importance of the Information Environment in supporting military operations. Activities in the physical domain are processed in the virtual domain of computers and networks before being interpreted by the human cognitive domain where decisions are made. Commanders at all levels have an ever increasing amount of information of varying latency and reliability available to them from a variety of sources. These range from highly sophisticated and complex bespoke surveillance systems to individuals equipped with a smartphone and Internet connection. An effective commander must be able to assimilate all the information sources available to them and be able to visualise the battlespace in an accurate and timely manner. To assist them, there are already a wide variety of software applications capable of receiving multiple inputs and displaying the disposition of military forces within a mapping environment. However, these tools place the interpretation and analysis responsibilities with the human operator. This paper proposes an AI driven process in which the initial analysis and correlation function is conducted in real time and in response to data inputs from multiple sources. This presents the decision maker with a fused, correlated and predictive Common Operational Picture providing a clear information advantage.