Effects of Automation in the Aircraft Cockpit Environment Skill Degradation Situation Awareness Workload Good Group #1: Julian Archer Hambisa Keno Yul Kwon

Automation Overview What is Automation? Types of Automation Automation is any system component that removes the necessity for direct human control of certain processes (Sherman, 1997) Types of Automation Control Warnings and Alerts Information Work on remember a simpler def Control – Replaces human control with machine control Warnings and Alerts – Replaces need for manual monitoring for hazardous conditions Information – Provides information that would otherwise have to be retrieved/ calculated manually

Automation Overview Early Stage 1914 The autopilot connected a gyroscopic Heading indicator and attitude indicator to hydraulically operated elevators and rudder It permitted the aircraft to fly straight and level on a compass course without a pilot's attention

Automation Overview Major Transition More control automation More alert and warning automation

Automation Overview Current

Automation Overview Benefits of Automation? Reduces workload and fatigue Precision in the handling of routine tasks Reliability Increased efficiency and productivity Economical utilization of machine Not always a good thing Like with another piece of technology that has an advantage, there is a disadvantage Based on frequency occurrence in the literature, three issues seemed initially interesting to look at…

I'm planning, don't disturb me! Skill Degradation Overview Inherent disadvantage of manual flying skill decay through non-practice Manual Flying Skills 1. Psychomotor skills Observable e.g. physical actions Psychomotor skills - Performed by the pilot on the controls to orient the aircraft Cognitive skills - Performed by the pilot to respectively monitor, plan and predict current and future states of the aircraft to satisfy required navigation objectives I'm planning, don't disturb me! 2. Cognitive skills Hidden e.g. mental actions

J. M. Childs, Spears, & Prophet, 1983 Skill Degradation Psychomotor Skill Degradation Performance decay is present regardless of time-lapse between training Performance quickly returns to a proficiency of 75% in 5 minutes of practice (Ammons et al., 1958) J. M. Childs, Spears, & Prophet, 1983

The series model of pilot control (adapted from McRuer, 1982) Skill Degradation Cognitive Skill Degradation Cognitive processing is a crucial skill involved in every aspect of flight Mental models help! The series model of pilot control (adapted from McRuer, 1982) Flight data combined with visual and sensory cues have to be synthesized for the pilot to make an appropriate and adequate control input to the aircraft The pilot must close several loops synchronously The ‘inner’ attitude control loop is closed in order to satisfy the ‘outer’ flight path control loop Control system limits the pilot and  allows for direct manipulation of the aircraft’s basic six degrees of freedom, i.e. translational movements (heaving, swaying, surging) and rotational movements (pitching, yawing, and rolling).   The pilot must then manipulate lower order parameters in order to control higher order goals (e.g. altitude and flight path) This control process is further complicated when lag is added to the system Research has shown that pilots who had significant experience flying non-glass cockpit aircraft developed robust mental models of performance characteristics during different phases of flight Over-reliance of the automation inhibits the pilot’s ability to develop robust mental models for manually aircraft control (Ebbatson, 2009)

(Adapted from Research Integrations Inc., 1997-2007) Skill Degradation General Evidence for Skill Degradation Accidents Investigations Studies Observations (Adapted from Research Integrations Inc., 1997-2007)

Skill Degradation Accident Analysis Data Skill-Based Errors Visual scanning breakdowns Poor technique Over-controlling the aircraft NTSB Accident records Commercial Aviation 1990 -1996 63.6% of aircraft accidents ≥ 1 skill-based error Constant over 7-year period Shappel & Wiegmann, (2000) Work on the other accident support

Skill Degradation Case study of Colgan Air Flight 3407 February 12th, 2009 Continental Connection Bombardier Dash-8 Q400 https://www.youtube.com/watch?v=sLMPFJQTmZs 28:37 to 29:33

(Adapted from Research Integrations Inc., 1997-2007) Situation Awareness Why Situation Awareness The most frequent causal factor of all accidents (41 percent) was lack of positional awareness in the air. (UK CAA Global Fatal Accident Review 1980 - 2007) The second most common primary causal factor was “lack of positional awareness in the air,” generally resulting in controlled flight into terrain (Flight Safety Digest December 2004–March 2005. Special FSF Report: Killers in Aviation) (Adapted from Research Integrations Inc., 1997-2007) From the statistics and reports

Situation Awareness What is Situation Awareness? “…the perception of the elements in the environment within a volume of time and space, the comprehension of their meaning, and the projection of their status in the near future”. (Mica Endsley, 1988) Even though correct or an absolute In terms of aviation

Situation Awareness Levels of Situation Awareness Seek and combine data into meaningful information (Level 1 : perception) Understand what the information means (Level 2 : comprehension) Use your understanding to think ahead and reconsider the plan (Level 3 : projection) Thinking ahead Updating the model 3 Understanding Comparison with mental models 2 Collecting and combining the data around you In this stage with the in u gathered ud try to understand actually means Where you take that meaning info to and use that to look ahead and prepare for the future. AS a pilot Anticipate the contingencies during the whole phase Scanning Gathering data 1 Feedback, check, monitor

Situation Awareness What Factors Reduce Situation Awareness? Workload Degraded operating conditions Pattern Recognition Crew issue Communication Attention I can explain a bit more in detail of what had happened since I was actually stationed at the base during the time of the accident. One of my friends who was in the investigation team told me that there were sequence of events that took place that resulted in reduction fo situation awareness. The flight crew had were coming in from australia that particular day. The weather was pretty bad with low visibility and Heavy tail wind. This factor made them do a circling approach and what it baisically is that you do your normal inst app and once you get to a Certain alt you would manuever to land on the other side of the rwy. The problem to this was even though they have flown over several tmes To the airport it was first time for both of them doing a cicling app at the airport. In the process the captain got mad at the co for not prepring For the flight the way he was supposed to. You can imagine the the high tension inside the cockpit. During that time the ATC acually steped in and Warned them about the hill front but therer was some confusion with the metrix they were using to indicate the altitude. Where the atc was using Ft as to the meter that the chinese crew were accustiomed to back home. This incident illustrates how multiple factors effecting SA can Lead to fatal accident like this one.

Situation Awareness Sometimes you see only half of the picture How does Automation Impact Loss of SA? Vigilance, Complacency and Monitoring Active vs. Passive Role Feedback Lack of understanding of automation Over relience on automation plays a big role. When the flight crew becomes accustomed in using automation and starts Trusting it too much it puts them in a more passive role rather than being more Actively involved in the flight process. Another factor would be insufficient feedback. Because todays automated systems are so complex the pilots have a hard time in Understanding what goes behind the scenes. Lack of understanding in automation can actually lead to An Increse in workload that can effect loss of SA. Hambi will talk more in detail of the subject. Sometimes you see only half of the picture but need all of it to understand the situation

Workload Humans have limited capacity for: Processing information. (display, alarm, communication, documentation) Holding items in memory Making decisions Performing tasks

Workload Excess Workload can result in human performance issues such as: Slower task performance Errors such as slips, lapses and mistakes Under-load can result in: Boredom Loss of situation awareness reduced alertness

Workload Measures of Workload Subjective Ratings Physiological Measures Performance measures

Workload Multi Attribute Task Battery (MATB) System Monitoring Tracking Communications Resource management

Workload The Irony of Automation Counterproductive workload consequences of automation where; Automation increases workload when already at its peak Lowers workload when pilots are under-loaded

Workload Role Transition Automation has resulted in transition of pilot role from controlling to monitoring Results in excessive increase in cognitive demand Humans are ill-suited for monitoring and will perform poorly

Workload Cognitive Overhead Performing extensive cognitive evaluation of the benefit of automation against cost of performing it manually

(Adapted from Research Integrations Inc., 1997-2007) Workload General Evidence for Workload Issues 82 documents addressed the impact of automation on workload 42 of these documents asserted automation can have adverse effect on pilot workload (Adapted from Research Integrations Inc., 1997-2007) transition

Conclusion Accident Statistics vs.Phases of Flight Distribution of fatal accidents by flight phase for 1998-2007 (Boeing, 2008)

Conclusion Addressing Skill Degradation Manual handling recency Challenges Operating procedures require automation Pilots flying Internationally are conscious of their lack of flying proficiency Rely on automation Pilots are mission-oriented Take Control Flip the Switch modify

Conclusion Augmented Displays Display that improves upon reality by superimposing info over actual environment The have been substantial improvements that are being made in effort to increase the crews SA Augmented heads-up displays and tech like enhanced vision system is making it Easier for the crew in obtaining more accurate info of its surroundings.

Conclusion Automations to improve SA Design improvements - Display design capitalizing on spatial relationships People in the earlier days were mostly focused on improving the tech itself But now they are actually working with the human beings from the developmental stages to look into how the System works in terms of addressing the human factor issues. This is actually much harder than in seems because The human factor issues are correlated in a very complex way. You might think more automation would decrease workload On the crew but in can be the other way around.

Conclusion Addressing Workload Issues Automation should allow for maintenance of optimal workload level (neither too high nor too low) Assessment of workload under automation should consider both steady state and transient operating conditions Task redistribution between team members offsets workload; automation driven single pilot operation concept needs to be approached with caution

Conclusion Addressing Workload Issues Automation workload assessment should consider visual and auditory input, cognitive activity and psychomotor skills The 4D resource theory can be used as a design input while developing human-machine interfaces to reduce pilot workload

Conclusion Task Allocation Strategies Design Time (Static)     Conclusion Task Allocation Strategies Design Time (Static) System designer sets the level of automation. Adaptive Automation Contextual allocation based on performance tracking Adaptable Automation Automation is a subordinate that collaborates with the human.

Conclusion Tasking Interface for Adaptable Automation User Interface Analysis and Planning Component

Conclusion Proof of Concept for Adaptive Automation http://www.sift.net/demos/playbook

Questions? http://www.sift.net/demos/playbook