Data Link and Technology
Integration Benefits to NAS
Performance
Jasenka Rakas
Wanjira Jirajaruporn, Tanja Bolic, Helen Yin
University of California at Berkeley
January 2006
1
Outline
‰ Issues
‰ Background
‰ Methodology
‰ Results
‰ Summary and Recommendations
2
Issues
‰ URET and Data Link Communications
Integration Benefits
‰ Data Link Benefit Analysis
3
Background
‰
‰
‰
‰
‰
Data Link Communications
Benefits assessment
User Request Evaluation Tool (URET)
Benefits of Data Link-URET integration
Excess Distance Analysis
4
Data Link
Data Link Communications is the Aeronautical
Data Link System (ADLS) that provides data
communications between aircraft and ground
automation system in the en route sector.
The Controller Pilot Data Link System
complements voice communications and
provides a link that is only used (in the initial
stages, i.e., in the Build I phase) for routine
messages, which make up about a half of all
controller/pilot communications messages.
5
High Sectors (FL240 to FL310 Approx)
Indianapolis/Memphis
Center Boundary
Indianapolis
Center (ZID)
Memphis Center (ZME)
6
Super High Sectors (FL330 and Above Approx)
Indianapolis/Memphis
Center Boundary
Indianapolis
Center (ZID)
Memphis Center (ZME)
7
Data Link
8
Data Link
‰ Build I
‰Transfer of communication
‰ Build IA
‰Assignments of speed, heading, altitude
‰Pilot initiated requests
‰Non-time critical messages from controller to pilot
‰ Build II
‰Enhancements, with emphasis on integration with
DSTs
9
Benefits and Benefits Assessment
‰ Reduced frequency congestion, especially
under high traffic density
‰ Benefits assessed as a reduction in
frequency occupancy depending on the Data
Link build in question
‰ Human factors should be looked at as well
10
URET
The User Request Evaluation Tool (URET) is the en route
controller tool used:
• to automatically detect and solve aircraft-to-aircraft and
aircraft-to-airspace conflicts
• for trial planning of proposed Flight Plan amendments
• for automated controller coordination of problem
resolutions
• for enhanced flight data management
The benefits and performance of URET have continuously
been evaluated since 1997.
11
URET
12
Data Link-URET Integration
‰
Benefits often assessed separately which can underestimate them
‰
Is it possible to send some conflict resolution messages via data link
13
Methodology
‰ Two parts:
‰Transcription and analysis of voice
communication
‰ Analysis of voice communication and traffic
“picture” under URET to identify possible
integration benefits
14
Transcribed Messages: Sample
Message
No.
Time
Headway
Type
A/C
Message Text
1
18:44:41
2
18:44:48
3
REQ
N34H
214
'32 to 39' '' ''
7
REM
N34H
214
'' ' indi center roger' ''
18:44:53
5
REM
UAL1235
447
'' ' contact kansas city center at 133.22' ''
4
18:44:55
2
REQ
UAL1235
447
'33 22 k c so long' '' ''
5
18:45:00
5
REM
RJX6160
243
'' ' contact indi center 134.27' ''
6
18:45:03
3
REQ
RJX6160
243
'34 27 good day' '' ''
7
18:45:16
13
RTE
UAL214
766
F ____ '' ' charleston' ' cleared present position direct to ____'
8
18:45:20
4
REQ
UAL214
766
'directly to charleston' '' ''
9
18:46:40
80
REQ
N92WG
988
'31 7 for 350' '' ''
10
18:46:44
4
REM
N92WG
988
'' ' indi center roger' ''
11
18:46:57
13
REM
N34H
214
'' ' contact memphis center 124.12' ''
12
18:47:02
5
REQ
N34H
214
'24 12' '' ''
13
18:50:20
198
REM
NWA286
333
'' ' contact indi center 134.17' ''
14
18:50:22
2
REQ
NWA286
333
'34 17 good day' '' ''
15
18:51:10
48
REM
UAL8170
567
'' ' contact indi center 134.27 34 27' ''
16
18:51:15
5
REQ
UAL8170
567
'34 27 good day' '' ''
17
18:51:26
11
REQ
UAL1940
355
you don't happen to have at 330 the wind and the ride do ya?' '' ''
18
18:51:36
10
REQ
SWA266
756
'370 good afternoon' '' ''
19
18:51:46
10
REM
SWA266
756
'' ' i appreciate it thanks' ''
20
18:51:51
5
REQ
TWA520
898
'crossing flight level 300 for 310' '' ''
21
18:51:57
6
ALT
TWA520
898
330 N '' '' C ' indi center' ' climb and maintain FL 330'
22
18:52:00
3
REQ
TWA520
898
'330' '' ''
15
RECORDER
TRANSCRIBER
Methodology
Excel file with transcribed
ATC messages
Analysis and selection of
ATC messages that ‘looked’
like conflicts.
MOPs:
1. time pilot checks into
the system (t2)
2. time controller issued
a clearance (t3)
3. message type
4. message duration
5. number of messages
6. aircraft type
7. frequency utilization
(congestion)
Data from the
Host Computer
URET
Identification of conflict
situations corresponding to
ATC transcribed messages
that ‘looked’ like conflicts.
XEVAL
Analysis of aircraft
conflicts, changes in
route amendments.
MOPs:
1. URET earliest alert time (t1)
2. URET
- predicted conflict
start time (t4)
3. total number of controlled
aircraft in sector per unit
time
4. number of URET alerts
MOPs:
1. aircraft conflict
profile (horizontal or
vertical)
2. number of aircraft in
conflict (two or more)
Verify conflict
MOPs:
A = t3 - t1 = time between controller’s clearance and URET’s first conflict notification
- predicted conflict start time and controller’s clearance
B = t4 - t3 = time between URET
C = t3 - t2 = time between controller’s clearance and the first time aircraft checks in.
16
Voice Communication Analysis
‰ Message analysis:
‰Type
‰i.e. initial call, handoff message
‰Different message types will be covered in different
CPDLC builds
‰Complexity
‰Frequency occupancy and message
duration
17
Voice Analysis Results
Number of Messages
%(total)
Pilots
605
50.17
Controllers
601
49.83
Total
1206
100.00
18
Voice Analysis Results
Pilots’ Message Type
No. of
Messages
%(pilot)
%(total)
Initial call (I)
158
26.12
13.10
Request (Q)
10
1.65
0.83
Acknowledgement (A)
348
57.52
28.86
Other (O)
89
14.71
7.38
Total (pilot)
605
100.00
50.17
19
Voice Analysis Results
Controllers' messages
by TYPE
No. of Messages
%
(controller)
%
(total)
Heading (H)
54
8.99
4.48
Altitude (L)
44
7.32
3.65
Fix (F)
42
6.99
3.48
Route Amendment (R )
12
2.00
1.00
Routine Message (M)
292
48.59
24.21
Other (W)
157
26.12
13.02
Total (controller)
601
100.00
49.83
20
URET- Data Link Integration Analysis
‰ Metric definition
‰ Identify clearances used for resolution of
URET conflict alerts
‰ Identify the clearances that could be sent via
data link
21
Question
Are there any clearances given to pilots
to resolve traffic conflicts
a sufficient time before the start
that we might consider delivering by data link?
22
Question
Is it feasible to communicate
conflict resolution messages using data
link in the integrated URET/Data Link
environment?
23
Question
If it is feasible,
what would be the impact on
frequency utilization?
24
Metric Definition: Relevant Times
t1
t2
t3
t4
t1 = time of URET conflict alert, which will occur at t4
t2 = time of initial call
t3 = controller issued a conflict resolution clearance
t4 = URET-predicted conflict start time
25
Metric Definition
C
B
A
t1
t2
t3
t4
A = t3 - t1 = time between controller’s clearance and
URET’s conflict notification
B = t4 - t3 = time between URET-predicted conflict start time
and controller’s clearance
C = t3 - t2 = time between controller’s clearance and the first time
aircraft checks in.
26
Description of the Study Area
Indianapolis Center
Super High-Altitude Sector 91
Memphis Center
27
URET Conflict
Display for
Three Aircraft
28
XEVAL
Control Panel
29
XEVAL Vertical
Conflict Analysis
for Three Aircraft
30
12
900
8
600
6
400
4
300
0
0
22:
45
30
15
22:
22:
00
30
45
22:
21:
21:
15
00
21:
21:
45
30
15
20:
20:
20:
00
45
20:
19:
30
15
19:
19:
00
45
19:
18:
30
15
18:
18:
2 :4
0
3 :0
5
0
0
2 :3
2 :0
5
5
1 :4
0
2 :1
-2
-2
-2
-2
1 :3
5
-2
1 :1
0
-2
-2
1 :0
0 :3
0 :4
0
5
-2
-2
-2
-2
9 :3
5
9 :4
0 :0
0
5
0
5
0:1
9 :1
0
-2
9 :0
5
-2
8 :4
0
-1
-1
-1
8 :3
-1
-1
-1
Aircraft Count
Average C (sec)
Average B (sec)
Average A (sec)
Aircraft Count
500
Time (seconds)
Comparison of Relevant Time Intervals Durations with Aircraft Count
800
10
700
200
2
100
Time (hours)
31
Comparison of Number of Messages with the Number of Aircraft Present, Potential Conflicts and URET Alerts
140
40
120
30
100
25
80
20
60
15
40
10
5
20
0
0
:4
22
:3
22
:1
22
:0
22
:4
21
:3
21
:1
21
:0
21
:4
20
:3
20
:1
20
:0
20
:4
19
:3
19
:1
19
:0
19
:4
18
:3
18
:1
18
5
0
5
0
5
0
5
0
5
0
5
0
5
0
5
0
5
0
5
-2
-2
-2
-2
-2
-2
-2
-2
-2
-2
-2
-2
-2
-1
-1
-1
-1
-1
-1
3:
2:
2:
2:
2:
1:
1:
1:
1:
0:
0:
0:
0:
9:
9:
9:
9:
8:
8:
00
45
30
15
00
45
30
15
00
45
30
15
00
45
30
15
00
45
30
Time (hours)
Aircraft Count
Number of Conflicts
Number of URET Alerts
Number of Messages
32
Messages
Aircraft, Conflicts and Alerts
35
Correlation
Number of
Aircraft In
Sector
Number of
Aircraft in
Sector
1
Number of
Conflicts
0.581
1
Number of
Messages
0.669
0.666
1
Number of
URET
Alerts
0.527
0.745
0.574
Number of
Conflicts
Number of
Messages
Number of
URET
Alerts
1
33
18
:4
5:
18 00
:5
5: 18:
0
50
0
19
- 1 :00
:0
5:
9:
00
19 00
- 1 :00
:1
5:
9:
10
19 00
- 1 :00
:2
5:
9:
20
19 00
- 1 :00
:3
5:
9:
30
19 00
- 1 :00
:4
5:
9:
40
19 00
- 1 :00
:5
5:
9:
50
20 00
- 2 :00
:0
5:
0:
00
20 00
:0
:1
5: 20: 0
10
20 00
- 2 :00
:2
5:
0:
20
20 00
- 2 :00
:3
5:
0:
30
20 00
- 2 :00
:4
5:
0:
40
20 00
- 2 :00
:5
5:
0:
50
21 00
- 2 :00
:0
5:
1:
00
21 00
- 2 :00
:1
5:
1:
10
21 00
- 2 :00
:2
5:
1:
20
21 00
- 2 :00
:3
5:
1:
30
21 00
- 2 :00
:4
5:
1:
40
21 00
- 2 :00
:5
5:
1:
50
22 00
- 2 :00
:0
5:
2:
00
22 00
- 2 :00
:1
5:
2:
10
22 00
- 2 :00
:2
5:
2:
20
22 00
- 2 :00
:3
5:
2:
30
22 00
- 2 :00
:4
5:
2:
40
22 00
- 2 :00
:5
5:
2
00 :50
- 2 :00
3:
00
:0
0
Frequency Utilization (%)
Frequency Utilization for Different CPDLC Builds
60
All messages
50
No Routine Messages
No Routine and Conflict Resolution Messages
40
30
20
10
0
Time (hours)
34
Conclusions
‰ There are significant benefits in frequency reduction
‰ Largest benefits during busy hours
‰ Possibility of lowering of communication errors
35
Data Link Efficiency Analysis
Objective
‰ Correlate Frequency Congestion and
Aircraft Excess Distance
36
Example of Potential Benefits
Efficiency Loss
Through UnTimely
Communication
CPDLC
Efficiency
Gain
Ac
f f ic
for Tra
n
o
i
t
a
Devi
Flight Plan Ro
ute
Optim
um S
e
CPD
L
tu
a
CR
out
parat
ed Ro
u
te
e
lF
low
nR
ou
te
CPDLC Efficiency Gain =
Actual Flown Route – CPDLC Route
Source: FFP1 Metrics Team
37
Example of Potential Benefits
u
tim
p
O
m
m
Cli
bP
ile
rof
tim
Op
ile
le
rof
P
of i
le
b
r
i
f
m
o
bP
Cli
Pr
m
i
d
b
l
te
C
lim
al
ara
C
u
p
t
C
Se
Ac
DL
P
um
C
Efficiency Loss Through
Un-Timely Communication
CPDLC Vertical Efficiency Benefit = Fuel Burn for Actual Climb Profile –
Fuel Burn for CPDLC Climb Profile
Source: FFP1 Metrics Team
38
Conceptual Framework: URET Sectors
Voice Tapes from MITRE for
ZID and ZME
Methodology
Data from the
Host Computer
Interactive Transcriber
URET
XEVAL
Excel file with transcribed
ATC messages
Identification of conflict
situations corresponding to
ATC transcribed messages
that ‘looked’ like conflicts.
Analysis of aircraft
conflicts, changes in
route amendments.
MOPs:
1. total number of
controlled aircraft in
sector per unit time
2. number of URET alerts
3.traffic congestion level
4. lateral plan
5. vectoring profile
MOPs:
1. aircraft conflict profile
(horizontal or vertical)
2. number of aircraft in
conflict (two or more)
Analysis of messages
MOPs:
1. message type
2. message duration
3. number of messages
4. number of conflicts
5. aircraft type
6. frequency utilization
(congestion)
MOP:
Route efficiency = F (voice congestion, aircraft conflicts, traffic congestion)
39
Example
•
Sector 91, ZID Center
URET traffic was analyzed and
conflicts were detected.
Xeval was used for detailed
analysis of conflicts/vectored
aircraft.
40
Example
From voice tapes and URET
planned route
A/C
ID
Conflict
with
aircraft
(from
URET)
Time
message
issued
(from
voice
tapes)
Controller’s
Message
(from voice
tapes)
NWA
574
AWE
240
19:25:32
“turn right 20
degrees please
for traffic”
NWA
574
AWE
240
19:29:25
“turn left 10
degrees”
NWA
574
AWE
240
19:30:
04
“you are
cleared left turn
now on course”
actual route
conflict
41
Example: Aircraft NWA574
Excess Distance from Xeval:
planned route
actual route
67
.5
NM
75
NM
excess distance = 7.5 NM
42
Relevant Information
TIME INTERVAL: 19:15 – 19:30 PM
Number of Communication Messages:
98
Aircraft in Sector 91:
8
Aircraft URET Alerts:
18
Conflicts:
13
Frequency Utilization:
19:15-19:20
19:20-19:25
19:25-19:30
19:30-19:35
20%
17.61%
37.57%
40%
43
List of Analyzed Sectors:
Indianapolis ZID Center
Sector
Time Interval (ZULU)
Sector Size
Conflicts (#)
(derived from
URET)
Name
(#)
Altitude
92
SH
19:15 – 19:45
med/large
peak
92
SH
21:45 – 22:15
med/large
low
95
SH
18:45 – 19:15
large
peak
98
SH
22:30 – 23:00
med/large
desc. peak
80
H
18:45 – 19:15
small
first peak
83
H
21:15 – 21:45
medium
low
84
H
21:00 – 21:30
med/large
medium
85
H
20:30 - 21:00
large
low
87
H
19:30 – 20:00
med/small
med. peak
89
H
20:45 – 21:15
small
second peak
89
H
22:00 – 22:30
small
second peak
44
Representative sample was made up of
various sectors with different:
- altitudes (H and SH)
- time intervals,
- sector sizes
- traffic demands and
- conflict types and counts.
45
Frequency utilization was calculated for 30-minute
segments, for previously listed sectors. Some
sectors included more than one sample.
Drop Page Fields Here
ZID 92, TIME: 19:30-20:00 ZULU
Total
Sum of Message duration(seconds)
100
90
80
70
60
Drop Series Fields Here
50
Total
40
frequency utilizationin in 5-minute intervals (%time)
30
20
35
10
30
25
0
19:25:00-19:30:00
19:30:00-19:35:00
19:35:00-19:40:00
19:40:00-19:45:00
19:45:00-19:50:00
5 minute interval
20
15
10
5
frequency utilizationin
5-minute intervals
(%time)
0
:1
5
19 :00
:2 -19
0
:
19 :00 20:
:2 -19 00
5
:
19 :00 25:
:3 -19 00
0
:
19 :00 30:
:3 -19 00
5
:
19 :00 35:
:4 -19 00
0
:
19 :00 40:
:4 -19 00
5:
00 : 45
-1 :00
9:
50
:0
0
19:20:00-19:25:00
19
19:15:00-19:20:00
46
Sector 80, Time period: 18:4519:15 (ZULU)
Frequency utilization in 5-minuite intervals (% time)
60
50
40
30
20
10
Frequency utilization in
5-minuite intervals (%
time)
Sector 98, Time period: 22:30 – 23:00
(ZULU)
0
18
:4
5
18 :00
:5 -18
0
18 :00 :50
:5 -1 :00
8
5
19 :00 :55
:0 -19 :00
0
19 :00 :00
:0 -1 :00
9
5
19 :00 :05
:1 -1 :00
9
0
19 :00 :10
:0
:1
5: 19: 0
00 15
-1 :0
9: 0
20
:0
0
Frequency utilization in 5-minute intervals (% time)
Frequency utilization in
5-minute intervals (%
time)
22
:3
0
22 :00
:3 -2
2
5
22 :00 :35
:4 -22 :00
0
22 :00 :40
:4 -2 :00
2
5
22 :00 :45
:5 -2 :00
2
0
22 :00 :50
:0
:5
5: 2 2: 0
23 00 55
:0 -2 :00
0: 3:
00 00
-2 :0
3: 0
05
:0
0
50
45
40
35
30
25
20
15
10
5
0
47
High and Super-High Sectors
‰ Used URET XEVAL tool to
analyze conflicts and determine
excess distance
‰ Excess distance increases by
0.076 nmi for each percent
voice channel occupancy
‰ Example: if CPDLC reduces
voice channel occupancy by 20
percentage points, expect 1.4
miles in savings per laterally
resolved conflict
‰ 50% of conflicts resolved
laterally
Sector
Time Interval (ZULU)
Sector Size
Conflicts (#)
(derived from
URET)
SH
19:15 – 19:45
med/large
peak
SH
21:45 – 22:15
med/large
low
95
SH
18:45 – 19:15
large
peak
98
SH
22:30 – 23:00
med/large
desc. peak
80
H
18:45 – 19:15
small
first peak
83
H
21:15 – 21:45
medium
low
84
H
21:00 – 21:30
med/large
medium
85
H
20:30 - 21:00
large
low
87
H
19:30 – 20:00
med/small
med. peak
89
H
20:45 – 21:15
small
second peak
89
H
22:00 – 22:30
small
second peak
Name
(#)
Altitude
92
92
High and Superhigh Sectors
12
Excess Distance
(nmi)
‰ Transcribed a total of ~18 hours
of voice communications
y = 0.0762x + 1.4182
10
8
6
4
2
0
0
10
20
30
40
50
Voice Channel Occupancy (% )
48
Summary
‰Engineering approach for solving complex
Data Link issues was presented.
‰Methodology using URET, Xeval and
Transcriber proved to be very useful and
well-designed.
‰Future work should include HUMAN
FACTORS ISSUES:
‰ Controller’s and pilots workload
‰ Situation Awareness
‰ Change in workload distribution from aural to primarily visual
49