Influence of task load on situation awareness and control strategy in the ATC tower environment
1. caso: ATC control strategy
1.1. Justification: Double air traffic estimated to 2020 in Europe, new systems without proper knowledge of related effects, increases probability of error
1.2. Task characteristics:
1.2.1. Time pressure
1.2.2. Multiple goals
1.2.3. Interconnected tasks
1.2.4. High consequences of error
1.2.5. Dynamic environment
1.2.6. Multiple information fonts
1.2.6.1. Navigation
1.2.6.2. Communication
1.2.6.3. Surveillance systems
1.3. ATC classification
1.3.1. En route control
1.3.2. Approach control
1.3.3. Tower control ATCOs
1.3.3.1. Use:
1.3.3.1.1. Air/ground communications
1.3.3.1.2. Radar
1.3.3.1.3. Direct view
1.3.3.2. Control:
1.3.3.2.1. Speed
1.3.3.2.2. Heading
1.3.3.2.3. Flight level
1.4. Objetive: understanding the influence of workload onto tower control operations for inroducing new technologies that increasing safety and efficiency.
1.4.1. How ATCOs interact with environment: whether probe techniques for SA are applicable for tower control operation and if is possible to measure the influences of increased task load on the control strategy
2. Human Performance adequate performances of Jobs, task and activities by operational personel
2.1. depens on the:
2.1.1. Capabilities, Skills, Knowledge, Motivation
2.2. Influenced by performance-shaping factors:
2.2.1. Communication, teamwork, trust, fatigue, stress, vigilance, attention, mental workload (measurement), situation awareness (measurement)
2.3. Workload, situation awareness and control strategies
2.3.1. Workload: when the difficulty of the task increases, more resources are required and the remaining capacity of the operator decreases
2.3.1.1. Task load: external stress or demand through task or system
2.3.1.2. Mental workload: resulting strain of these stresors on the person depending on their individual characteristics
2.3.1.3. High workload leads to poor performance adn higher number of errors
2.3.2. Situation awareness:
2.3.2.1. Ensley (1995): a hierachical construct comprising the perception of elementes in the environment together whit their states and atributes, the understanding of these elements and their significance, and projection of future states of the elements and the situation
2.3.2.2. Durso et al. (1998) implicit process that involve simply knowing where the environment to find a particular piece of information, rather tan remembering what piece of information is
2.3.2.3. Chiappe et al. (2012) indiviuals forming a detailed mental representation that spares the limited capacity of working memory and attention
2.3.2.4. SA used in high-risk environments like: medical operating room, submarine track management, or nuclear power plants
2.3.2.5. It's a necessary and mesurable construct for understanding human computer interactions workload
2.3.2.6. Approaches to measuring:
2.3.2.6.1. 1) Subjetives measures self-rating or experts (3D Situational Awareness Rating Technique or SASHA) "the people are not always aware of that they don't know
2.3.2.6.2. 2) Implicit performance measures (observable actions and errors) but isn't clear realtion between SA and performace
2.3.2.6.3. 3) Probe technique, asking subject questions about current situation
2.3.3. Control Strategies: there are several possible control strategies for a certain traffic situation that differ in safety, efficiency or fuel economy and thus lead to different performance. (Tolerance for variability of situations main factor)
2.3.3.1. Papenfuss and Friedrich (2016) used eye trecking to identify the different control strategies
2.3.4. Relation with performance:
2.3.4.1. Good SA can reduce the processing capacity of future decisions and thus time pressure
2.3.4.2. Requires capacity for its formation in the time span until a decision is made
2.3.4.3. Current state of research indicates that none of the construct (workload, SA, control strategy and performance) can be measured and analyzed on its own.
2.3.4.4. Dynamic environmet inluence workload and selected contol strategy, they influence each other directly, whereas SA can be affected if the strategy demands change
3. Results
3.1. performance
3.1.1. Safety-critical performances separation los, happened twice and was caused by pseudopilots
3.1.2. Number of takeoffs Scenario A M=8 an SD=1,32, Scenario B M=8,67 SD=0,71
3.1.3. Taxi time: A= 260-269, B= 238-266
3.1.4. Duration radio communication A= 4,15-5,04, B = 3,96-4,45
3.2. Effects on SA
3.2.1. SARA-T (subjective)= correlation between SA, age & Experience (time response)
3.2.2. SASHA (probe) = scores not differ significantly for each scenario.SA higher under low visibility conditions
3.2.3. The majority of the participants deemed the SARA-T questions moderately not appropiate for assesing their situation awareness. questions were more distracting in phases with more than four aircraft
3.3. Effects on control strategy
3.3.1. Dwell time: AoI flight-stips (40%), Radar (20%), video panorama (17%), aircraft (11%), Weather and time (5%), SARA-T (6%)& radio (1%), results for both scenaries do not differ.
3.3.2. Connection between SASHA and SARA-T. If the reaction time increase the SASHA scores indicate less situation awareness.
3.3.3. If the task load is high, more visual attention has to be directed to the radar and less to the aircraft
3.3.4. Kendall's tau test= only the reponse time within the high task load phases show a significant correlation with the SASHA score.
4. Discussion
4.1. The question therefore remains as to whether SA is a factor influencing performance or if SA is the outcome of a task, thus performance itself. This question cannot be answered.
4.2. The results showed the existence of different estrategies for ATC environment, one for high task load pases that concentrases on radar information, and another one for los tasa load phases that focused more on aircraft
4.3. The SASHA tendency for contradict the assumption that reducen visibility impairs SA was supported by the analysis of SARA-T
5. Method
5.1. Sample: 9 participants from tower Braun-schweig-Wolfsburg and via an internal German Aerospace Center. average age 44 years, from 3 to 33 years of expereince
5.2. Design: simulated ATCO position, 2x2 eye movement and performance under a) task load equal for both scenarios A y B 10 min low x3, 5 high x 2 b) different visual conditions low vs high visibility. Each scenario take 40 min (pseudopilots), max 8 aircraft.
5.3. Apparatus: headset and eye-tracking glasses in a simulator, SARA-T presented in a separate display
5.4. Performance measures:
5.4.1. Safety = # separation losses
5.4.2. Efficiency= # takeoffs, taxi time, duration radio communication
5.5. SA measurement:
5.5.1. Self-rating questionnaries SASHA (6 questions, 7 points scale "never to always" (table 3)
5.5.2. SARA-T questions covered perception as well as comprehension
5.6. Eye tracking: SMI Eye Tracking Glasses 2 to identify possible control strategies that interact with SA and task load
5.6.1. Eye direction
5.6.2. Gaze
5.6.3. Head position