Softwalls, E. A. Lee 1
Preventing the use of CommercialAircraft as Weapons
Edward A. LeeProfessor, EECS, UC Berkeley
with Xiaojun Liu, Adam Cataldo
Institute of Transportation Studies SeminarFebruary 8, 2002
Softwalls, E. A. Lee 2
A Lethal Weapon?
Softwalls, E. A. Lee 3
Need to Shield
• Major cities• Government centers• Chemical and nuclear plants• Military installations• Critical infrastructure
Softwalls, E. A. Lee 4
Softwalls
• Carry on-board a 3-D database with“no-fly-zones”
• Enforce no-fly zones using on-boardavionics (aviation electronics)
• Non-networked, non-hackable
Softwalls, E. A. Lee 5
Normal responsepilot steeringcontrol.
Biasing Pilot ControlWhile Maintaining Responsivity
pilot'sdesireddθ/dt
actualdθ/dt
Right
Right Left
M
M− M
nobias
− M
Left
θ = heading angledθ/dt = rate of change of headingM = maximum rate of change
Softwalls, E. A. Lee 6
Bias the aircraftto the right.
Biasing Pilot ControlWhile Maintaining Responsivity
pilot'sdesireddθ/dt
actualdθ/dt
Right
Right Left
M
M− M
nobias
− M
Left
θ = heading angledθ/dt = rate of change of headingM = maximum rate of change
Softwalls, E. A. Lee 7
Biasing Pilot ControlWhile Maintaining Responsivity
Bias of –M allowsthe plane to turnonly to the right.
pilot'sdesireddθ/dt
actualdθ/dt
Right
Right Left
M
M− M
− M
Left
bias = − M
θ = heading angledθ/dt = rate of change of headingM = maximum rate of change
Softwalls, E. A. Lee 8
Biasing Pilot ControlWhile Maintaining Responsivity
Bias of –3M/2forces the planeto turn to theright at least athalf the maximumrate of change ofheading.
pilot'sdesireddθ/dt
actualdθ/dt
Right
Right Left
M
M− M
− M/2
M/2
bias =− 3M/2
− M
Left
θ = heading angledθ/dt = rate of change of headingM = maximum rate of change
Softwalls, E. A. Lee 9
Objective is toallow the pilotmaximum latitudesubject to no-flyzone restriction
Responsivity
Responsivity of 1ensures thataircraft dynamicsappear normal,and the bias feelslike an externalforce.
pilot'sdesireddθ/dt
actualdθ/dt
Right
Right Left
M
M− M
− M
Left
responsivity = slope
θ = heading angledθ/dt = rate of change of headingM = maximum rate of change
Softwalls, E. A. Lee 10
Sailing Analogy – Weather Helm
force ofthe wind onthe sails
turnedrudderkeeps thetrajectorystraight
withstraightrudder
with turnedrudder
Even with weather helm, thecraft responds to fine-graincontrol as expected.
Softwalls, E. A. Lee 11
A Preliminary CandidateControl Strategy
due to Xiaojun Liu
Softwalls, E. A. Lee 12
Two-Dimensional Aircraft Model
• speed s• position p• heading θ• time t
&( ) ( ( ) cos( ( )), ( ) sin( ( )))p t s t t s t t= θ θ
Softwalls, E. A. Lee 13
Criticality – Time to Wall
• Measure of time to wallin the worst case (mostuncooperative pilot)
• Assumes the pilot turnstoward thewall at themaximumrate
No-FlyZone
c = d/v
c = π/2/M+(d-rmin )/v
c = π/M
rmin
rmin
Softwalls, E. A. Lee 14
Maximally Uncooperative Pilot
• Assume θ = 0is headingtowards thewall
• This pilotsteersmaximallytowards thewall
θπ
−πT
− T
M
− M
pilotcontrol
Softwalls, E. A. Lee 15
Bias from Criticality-Based Controller
• If time to wall is less than π/M,the bias rises– at the wall, heading away is OK
• At 2/M it saturates.– still can avoid wall with half-
maximum turn.
bias
3M/2
π/M2/Mc(x, θ)
Softwalls, E. A. Lee 16
Simulation Model
aircraft model
criticality calculation
pilot model
bias control
Softwalls, E. A. Lee 17
Simulation – Maximally Uncooperative Pilot
Assumptions (pulled out of a hat):• speed: 0.1 miles/sec = 360 miles/hour• M: 2π/20 radians/sec• min turning radius: speed/M = 0.32 miles
pilot turns towards the wall
the
wall
bias starts, pilot counteracts
pilot controls saturate
pilot regains steeragetowards wall
nautical miles
Softwalls, E. A. Lee 18
Related Methods
• Ground proximity warning systems• Automatic ground avoidance systems• TCAS & ACAS – collision avoidance• Potential field methods for air-traffic control
HoneywellTCAS
Rockwell conflict resolution
Softwalls, E. A. Lee 19
Objections
• Reducing pilot control is dangerous– reduces ability to respond to emergencies
Softwalls, E. A. Lee 20
Is There Any Aircraft Emergency Severe Enoughto Justify Trying to Land on Fifth Ave?
Softwalls, E. A. Lee 21
Objections
• Reducing pilot control is dangerous– reduces ability to respond to emergencies
• There is no override– switch in the cockpit
Softwalls, E. A. Lee 22
No-Fly Zone with Harsher Enforcement
There is nooverride in thecockpit thatallows pilots tofly throughthis.
Softwalls, E. A. Lee 23
Objections
• Reducing pilot control is dangerous– reduces ability to respond to emergencies
• There is no override– switch in the cockpit
• Localization technology could fail– GPS can be jammed
Softwalls, E. A. Lee 24
Localization Issues
• GPS• Inertial navigation
“Localization” is the technology forreliably and accurately knowing thelocation of an object.
Softwalls, E. A. Lee 25
Objections
• Reducing pilot control is dangerous– reduces ability to respond to emergencies
• There is no override– switch in the cockpit
• Localization technology could fail– GPS can be jammed
• Deployment could be costly– how to retrofit older aircraft?
Softwalls, E. A. Lee 26
Deployment
• Fly-by-wire aircraft– a software change
• Older aircraft– autopilot level
• Phase in– prioritize airports
Softwalls, E. A. Lee 27
Objections
• Reducing pilot control is dangerous– reduces ability to respond to emergencies
• There is no override– switch in the cockpit
• Localization technology could fail– GPS can be jammed
• Deployment could be costly– how to retrofit older aircraft?
• Deployment could take too long– software certification
Softwalls, E. A. Lee 28
Not Like Air Traffic Control
This seems entirelyindependent of airtraffic control, andcould complementsafety methodsdeployed there.Self-contained on asingle aircraft.
Softwalls, E. A. Lee 29
Objections
• Reducing pilot control is dangerous– reduces ability to respond to emergencies
• There is no override– switch in the cockpit
• Localization technology could fail– GPS can be jammed
• Deployment could be costly– how to retrofit older aircraft?
• Deployment could take too long– software certification
• Fully automatic flight control is possible– throw a switch on the ground, take over plane
Softwalls, E. A. Lee 30
UAV Technology(Unoccupied Air Vehicle)
e.g. Global Hawk(Northrop Grumman)
Technology Support Working Group(TSWG), office of the Secretary ofDefense, has reportedly decidedagainst recommending any partialcontrol approach. Their feeling isthat there is only one feasiblestrategy: a single trigger, either on-board or remote control, that wouldassume complete control and takethe plane to a safe base.
Northrop Grumman has such asystem in the Global Hawk UAV thatsome believe can be dropped-in topassenger airliners.
Softwalls, E. A. Lee 31
Potential Problems with Ground Control
• Human-in-the-loop delay on the ground– authorization for takeover– delay recognizing the threat
• Security problem on the ground– hijacking from the ground?– takeover of entire fleet at once?– coup d’etat?
• Requires radio communication– hackable– jammable
Softwalls, E. A. Lee 32
Open Questions
• Technical issues– Geometry constraints on no-fly zones?– Can localization without GPS be accurate enough?– Can the database be secure?– Can areas near urban airports be protected?– How to prove safety?– Robustness with partial system failures?
• Policy issues– Definition of no-fly zones– Centralized vs. decentralized control
Softwalls, E. A. Lee 33
Acknowlegements
• Adam Cataldo• David Corman (Boeing)• Peter Huber (Forbes Magazine)• David Lee• Xiaojun Liu• Per Peterson• Shankar Sastry• Claire Thomlin• Don Winter (Boeing)• Paul Yang
Softwalls, E. A. Lee 34
Discussion
• Absent terrorism, does this make flying safer?• Is it better to have F-16’s enforcing no-fly zones?• Are pilots willing to give up some control?• Can the technique be phased in?• Are there other, simpler approaches?• …