A stochastic conflict detection model revisited

1
CoSpaceExperimental results on sequencing
mergingKarim Zeghal
ASAS Thematic Network, Second workshop 6-8,
October 2003, Malmö, Sweden
2
Introduction

• Motivation
• Increase of controller availability through a
  better allocation of spacing tasks between air
  and ground
• Neither to transfer problems nor to give more
  freedom to pilots!
• One option to improve safety, and beyond
  efficiency and/or capacity
• Constraints
• Human consider current roles and working methods
• System keep things as simple as possible
• Assumptions
• Airborne surveillance capabilities (ADS-B, state
  vectors)
• Airborne functions (ASAS, manual mode)

3
Principles

• Starting point
• Analogy with visual separation clearances
• but no transfer of separation responsibility
• Just new spacing instructions
• Spacing not separation, instruction not clearance
• To be used with current practices
• FAA/Eurocontrol PO-ASAS and ICAO SCRSP ASAS
  circular spacing category
• Task distribution
• Decision-making on the ground (controller defines
  strategy)
• Execution in the air (pilot implements actions)
• Two classes of operations
• Crossing and passing en-route
• Sequencing in terminal areas

4
Sequencing of arrival flows

• Controller
• Defines sequence order
• Issues spacing instruction
• Pilot
• Adjusts speed to acquire and maintain spacing
• Not authorised to change trajectory nor altitude

WPT
spacing
AFR123 235 ? 40
DLH456 250 ? 41
5
In situation
Controller AZA324, select target 2443 Pilot
AZA324, target 2443 identified, 3 oclock, 30
miles Controller AZA324, behind target merge
to INKAK to be 8 miles behind Controller
ASAS324, cancel spacing, reduce speed 220 knots
6
Real-time simulations
Joint EVP
CRZ-IAF interface
CRZ-IAF activity
CRZ-IAF activity/margins
Initial ideas IFATCA98
air
ETMA / ER exploratory
TMA exploratory
ground
ETMA / ER activity
ETMA monitoring control quality very high traffic
ETMA time/distance
Joint NUPII
1998
1999
2000
2001
2002
7
Controller E-TMA simulation
Joint EVP
CRZ-IAF interface
CRZ-IAF activity
CRZ-IAF activity/margins
Initial ideas IFATCA98
air
ETMA / ER exploratory
TMA exploratory
ground
ETMA / ER activity
ETMA monitoring control quality very high traffic
ETMA time/distance
Joint NUPII
1998
1999
2000
2001
2002
8
Experiment setup

• Overall
• Six controllers during 3 weeks
• Dense and generic airspace (simplified Paris
  South-East arrivals)
• Four (combined in two) arrivals sectors
• All traffic equipped
• Use of spacing at controller discretion
• No specific tools (paper strips with graphical
  markings only)
• Independent variables
• Spacing without, distance, time
• Sector configuration

Sequencing constraint 8Nm at IAF
9
(No Transcript)
10
Controller activity

• Natural mapping of the sequencing activity over
  the geographical sector

2- Maintaining
1- Building
Sequencing phases
Heading
Speed
Types of instruction
Converging point
Exit point
11
Spatial mapping of instructions
Heading, direct, speed (and spacing)
Very high Without
Very high With
12
Distribution of instructions (2001)
13
Monitoring

• Global level
• Reduced amount of time associated to monitoring?
• Local level
• Aircraft still monitored?

Use of eye movement analysis
Collaboration with NOVADIS, Grenoble
14
Eye-tracker data
5 seconds
5 minutes
15
Distribution of eye-fixations
High
25
25
High
Without
With
of fixations
20
20
15
15
Number of instructions
10
10
Distance to exit (Nm)
5
5
0
0
0
20
40
60
80
100
120
140
160
180
200
0
20
40
60
80
100
120
140
160
180
200
Very High
Very High
25
Without
With
20
15
10
5
0
0
20
40
60
80
100
120
140
160
180
200
0
20
40
60
80
100
120
140
160
180
200
16
Same trend in 2002
17
Spacing at exit point
100
90
90
80
71
70
59
60
47
45
50
40
30
20
16
20
13
6
10
2
2
1
0
Unacceptable - 5Nm - 60s
Small 5 - 7.5Nm 60 - 85s
Optimal 7.5 - 8.5Nm 85 - 95s
Large 8.5Nm 95s
Without
Distance
Time
18
Summary ground E-TMA

• Initial understanding of impact on controller
  activity and effectiveness in E-TMA
• Increased availability (instructions,
  eye-fixations)
• More stable flows transferred to the approach
• Seems to be beneficial when properly used
• Issues
• Applicability conditions (nothing more than
  todays practices) if not respected, use of
  spacing worse than conventional control
• Too much expectation? Risk of disengagement?
• Abnormal situations
• Applicability to other E-TMA airspace (even AO vs
  AR)
• Applicability (usefulness) in TMA

19
Controller TMA simulation
Joint EVP
CRZ-IAF interface
CRZ-IAF activity
CRZ-IAF activity/margins
Initial ideas IFATCA98
air
ETMA / ER exploratory
TMA exploratory
ground
ETMA / ER activity
ETMA monitoring control quality very high traffic
ETMA time/distance
Joint NUPII
1998
1999
2000
2001
2002
20
Understand TMA specificity

• Airborne spacing appropriate in E-TMA but
  problematic in TMA?
• Anticipation needed with spacing hardly
  compatible with todays practices (late vectors
  for final integration)?
• Difference between E-TMA and TMA?
• With E-TMA sectors previously considered and
  associated approach sectors (Paris Orly and CDG)
• Standard trajectories (E-TMA) vs radar vectoring
  (TMA)
• Integration on a point (E-TMA) vs on an axis
  (TMA)
• plus high time critically (pressure, lack of
  space, larger turns) generating uncertainty and
  preventing early planning

21
Experiment setup

• Objective assess usability of proposed
  applications
• Setup
• Four approach controllers during 9 days
• Standard trajectories, merging point, INIITM
  grouped, with EXC and PLC
• Two approach sectors
• Independent variables without spacing,
  distance, time
• All traffic equipped
• Use of spacing at controller discretion
• Traffic coming under spacing
• 31 per hour with sequence of up to 5

June02
22
Method of use

• From same IAF
• Kept in remain with direct
• Use of heading then merge (to create spacing)
• From different IAF
• Use of merge
• Use of heading then merge (to create spacing)

23
Distribution of instructions (without)

• Method of analysis used for E-TMA and translated
  to TMA

24
Distribution of instructions (with)
25
Distribution of instructions (all)
INIO
Without spacing
Distance based spacing
Time based spacing
INIR
26
Summary ground TMA

• Usability
• Seem usable in TMA under medium-high traffic
• Change in working method (standard trajectories,
  final integration on a point, unique approach
  control position)
• Impact
• Analysis of instructions suggests a positive
  impact on activity
• Potential for providing more availability
• Provides but also requires anticipation
• Allows to smooth traffic but gives the feeling
  of loosing capacity (less pressure)
• Issues
• Cost of cancelling then (re-)initiating spacing
• Reluctance to cancel spacing leads to group
  aircraft
• Recovery of abnormal situations

27
Flight deck simulation
Joint EVP
CRZ-IAF interface
CRZ-IAF activity
CRZ-IAF activity/margins
Initial ideas IFATCA98
air
ETMA / ER exploratory
TMA exploratory
ground
ETMA / ER activity
ETMA monitoring control quality very high traffic
ETMA time/distance
Joint NUPII
1998
1999
2000
2001
2002
28
Summary flight deck

• General
• Overall positive feedback on concept and
  interface
• Active part (in the loop) and understanding of
  the situation
• More anticipation and optimised flight
  management?
• However, a new task with potential risk of
  workload increase
• Managed (automatic) mode helpful
• Tolerance vs activity
• Tolerance at or above 0.5Nm seems acceptable
  (under nominal conditions and down to initial
  approach)
• Lower impact of tolerance than anticipated (keep
  the bug aligned culture)
• Issues
• Down to final approach (June 03)
• Abnormal situations

29
Next

• Ground (with NUPII Bretigny)
• 2003 TMA under very high traffic
• 2004 Interaction TMA / E-TMA with AMAN, and
  uplink for target selection, downlink of spacing
  parameter
• 2005 Extend scope towards assessing impact on
  ATC
• Air (with EVP WP3)
• 2004 From cruise to final approach, in varied
  conditions (long sequences) Full flight
  simulator (A330 from TuB/ZFB)
• 2005 Extend scope towards automatic mode and
  uplink for target selection