The Human Factors of Unpiloted Aircraft

1
First Conference on the Commercialization of
Unmanned Aviation
The Human Factors of Unpiloted Aircraft
All aspects of UA Control The Human Factors
2
Unmanned Aircraft (UA) Reconnaisance for Home
Use 259.00
November Issue of Mindware Catalog 2005
3
The qualifications and status of remotely
piloted vehicle operators are among the most
controversial aspects of remotely piloted vehicle
development Opinions about who should be
futureoperators range anywhere from the man off
the street to a highly qualified pilot with an
engineering background (Kiggans, 1975).
4

• Unpiloted Aircraft Unknowns
• Voluminous especially Control
• No Plan in Place TrialError
• It is happening Fast
• Order and Reason Needed

• Methodologies Available to Answer Unknowns
• WG, WAG, SWAG
• Carefully controlled lab and field studies
• Carefully controlled Human Operator simulations

5
Human Factors Systems Research Labs
Implement
6
UAVs

• Advantages
• Longer duration flights
• More radical flight maneuvers
• Smaller and lighter
• Stealthier
• Lower life-cycle costs
• More easily transported, and can
• prevent the endangerment of human life
• Current problems

7
Automation

• Automation doesnt always reduce workload
• MABA-MABA (men are better at-machines are better
  at)
• Humans susceptible to
• Vigilance decrements
• Monitoring complacency
• Fatigue
• Boredom

8
Endsleys Level of autonomy on Operator
performance and SA
Joystick Control
Voice Control
Scripts
Voice and Touchscreen
9
The Problem

• There are thousands of research variables that
  would need to be considered. It would take years
  to figure out which ones were the most important
  and decades to study.

• UA are appearing in the NAS much faster than that
  and the conditions must be resolved quickly.

• How can you accelerate the study of human
  factors?

10
A Solution Use modeling tools to study human
factors concerns

• Analyze the effects of crew size and fatigue on
  the operation of a TUA.

• Develop a simulation model to reflect mission
  parameters of a TUA operators task..

• 1st step Develop algorithms to show performance
  degradations from fatigue and time of day.

11
Human Factors Systems Research Labs
Implement
12
(No Transcript)
13
TUAV Simulation Model

• Data from 20 SMEs at Ft. Huachuca were used to
  build a model simulating TUAV missions under
  different types of searches, terrains
  emergencies.

14
Integrated Network of Operator Tasks Discrete
Event Simulations

• Modeled as series of individual tasks that are
  driven by spreadsheet data.
• Each crewmember is assigned schedules for each
  condition.

15
Variables
16
Tactical Operations Center Launch/Recovery Station
Rotation Schedules

• Rotation schedules are determined by Coverage,
  Crew Size (4,6 and 8 operators), and restrictions
  placed on how long soldiers can be on duty.

Surge was 18 hours of flight in 24 hours for 3
consecutive days
18 HR SURGE 2 hr rotation
17
Research Issue for this Approach Find
Appropriate Duty Cycle Crew Size Mix
Fatigue The Enemy ALIASES Exhaustion,
Weariness, Lassitude, Apathy, Languor, Burn out,
Malaise

• Intelligence Threat
• Errors in judgment
• in accuracy
• Decreased response time
• Irritability
• Blurred vision
• Muscular weakness

18
Precedents
Advanced Computer Generated Force Simulation
studies consider the impact of nuclear attack and
heat stress scenarios on combatant availability.
Why not consider fatigue other stressors to
find bottlenecks, particularly when there are
myriad what if questions ?
19
Fatigue Algorithm

• The Fatigue Degradation Equation (FADE)
  attempts to realistically degrade response time
  in models of human behavior.

• It is based on empirical data collected from
  USAF pilots during a 52 hr sleep deprivation
  cognitive testing period.

• The main focus of FADE is three
  fold the effects of prolonged
  sleep deprivation of
  circadian
  disruption of sleep recovery

20
Human Factors Systems Research Labs
Implement
21
SCAT (Space Cognitive Assessment Test)
Performance Degradation Over Time With FADE
Estimate of Performance
22
ALCOHOL ETOH
Dawson, D. and Reid, K. Fatigue, alcohol and
performance impairment. Nature, 388 235, 1997
23
Results (1 of 4)

• Almost 3 times as many AV mishaps can
  occur when a crew is fatigued than when a crew is
  well rested.

24
Results (2 of 4)

• As crew size decreased, the
  amount of time it took to search for a
  target increased.

25
Results (3 of 4)

• As crew size decreased, the
  number of targets that were detected
  decreased.

26
Conclusions

• Fatigue is arguably one of the most persistent
  threats to accelerated mission requirements. The
  12 hour surge may be excessive.

• Reducing the number of soldiers operating a TUAV
  can result in
• More potential AV mishaps during emergencies
• An increase in the time it takes to search for
  targets
• A decrease in the number of targets detected.

27
Why Modeling and Simulation Approach?

• Challenges
• Rapid virtual prototyping of the problem in human
  computer interface
• Dynamic, fluid and stressful conditions can be
  approximated
• Conduct relevant research to support the modeling
  efforts
• Thousands of experiments can be conducted
  rapidly.

• Benefits
• MS studies are statistically meaningful
• MS studies will allow the newest innovations to
  be tested
• MS studies will save time spent on research
  identifying the key variables for validation
  studies.

28
Human Factors Systems Research Labs
Implement
29
FIELD STUDIES EAGLEWORKS Engineering Human
Factors Research
30
Human Factors Systems Research Labs
Implement
31
ERAU can go from Simulation to Simulator to
Sensiferous
32
(No Transcript)
33
Human Factors Systems Research Labs
34
Multi-Modal Immersive Intelligent Interface for
Remote Operations (MIIIRO)
Dr Liu MIIIRO Lab Director Human Factors Embry
Riddle University
35
MIIIRO
Dr Liu MIIIRO Lab Director Human Factors Embry
Riddle University
36

• PROPOSE TO EVALUATE
• UA Operator duty cycles to find optimal shift
  schedule for operational demands.
• Incorporate fatigue into model
• Identify tasks that are repeated so often they
  might best be automated.
• Identify critical tasks that make workload
  excessive or create bottlenecks and evaluate
  countermeasures.

• Evaluate myriad other problems associated with UA
  operator tasking Examples
• Bandwidth
• Level of control
• Mode of control

37
Chelton on a Frasca Half-Shell
We train pilots in a simulated aircraft. We will
need to train UA operators in simulators. Soon
the line will blur even more between simulation
and reality.
Elite Simulator
38
Questions ?
UA are here! It is safer and cheaper to
rationally prepare, as much as possible, to
factor their implementation into the NAS, the
human world, than to fix problems as they occur -
Anon
39
Project had its origins in a project between DOD
and California Micorwave. They built 5 aircraft
which were modified Burt Rutan long ezs. I these
were optionally manned vehicles. That is if the
guy capturing the surveillance was taking fire,
he could land on a friendly road, get out and
then fly using his Toshiba hand held model
ac controller. Fly LOS back,take video etc then
have the a/c pick him up and off they go. Of the
5 aircraft, three were cut up at the end of the
contract, one crashed testing control laws and
the last made its way into the civilian
sector where it was flown as a single pilot veh
for some 10 years. We have that a/c here. I flew
it here. It has been gutted in preparation for
installing control and payload systems. We have
integrated a new Fabrique National 50 cal system
as well as a 2.75 inch launcher. See attached.
Let me know.
40
(No Transcript)
41
(No Transcript)
42
(No Transcript)
43
(No Transcript)