JPDO Systems Modeling and
Analysis
Dr. George Hunter
Presented on behalf of Yuri Gawdiak
JPDO Systems Modeling and Analysis Division
(SMAD)
September 5, 2007
Systems Modeling and Analysis Division
JPDO Organizational Changes
• Yuri Gawdiak has joined the JPDO as the lead for
the newly created Systems Modeling and Analysis
Division (SMAD).
• This division will continue the SEAD's work on
modeling the NextGen architecture and systems.
• SMAD will also provide analytical support to other
JPDO offices to include the Enterprise
Architecture and Engineering Division, the
Portfolio Management Division, and the Policy
Division.
2
Systems Modeling and Analysis Division
JPDO Organizational Changes
• Yuri comes to the JPDO from NASA and has
extensive experience in aeronautical engineering
and analysis.
• He managed the development and transfer of air
traffic management and safety applications to the
FAA and industry and was the program manager
of the Engineering for Complex Systems
program.
• Most recently, he has been performing strategic
analyses within NASA's Program Analysis and
Evaluation Office.
3
Systems Modeling and Analysis
Division
•
Priorities:
–
–
–
•
Better coordination with Enterprise Architecture & Portfolio
Management Divisions
Support improved Operational Improvement prioritizations
and sequencing
Develop and implement comprehensive verification &
validation strategy
Approach:
–
–
–
Collect requirements and schedule targets from JPDO
divisions
Conduct technical interchange meetings with partner
organizations & programs to get lessons learned,
expectations, issues, requirements, and identify possible
tools/products that are mutually beneficial
Develop integrated SMA division plan
4
SMAD Key Planning Elements
• Stakeholder Identification
• Requirements Analysis
– Fixes to existing functions
– Upgrades/performance improvements
– New functions/gap fillers
• Current key goals in the SMAD plan
– Improve turn-around time improvements/quick response
capability
– Integrated JPDO/FAA support schedule
– Strategic upgrades (including long term validation
approaches)
5
Systems Modeling and Analysis Division
JPDO Goals
Expand Capacity
Ensure our
National Defense
Ensure Safety
Protect the
Environment
Secure the Nation
Retain U.S. Leadership
in Global Aviation
6
Systems Modeling and Analysis Division
What is NextGen?
• Next Generation Air Transportation System
• The “end state” of the JPDO’s work (2025)
Operating Principles
Key Capabilities
• “It’s about the users…”
• Net-Enabled Information Access
• System-wide
transformation
• Performance-Based Services
• Prognostic approach to
safety assessment
• Weather-Assimilated Decision
Making
• Layered, Adaptive Security
• Globally harmonized
• Broad-Area Precision Navigation
• Environmentally
compatible to foster
continued growth
• Trajectory-Based Aircraft
Operations
• “Equivalent Visual” Operations
• “Super Density” Operations
7
Operational Improvements (OIs)
•
Each segment in the Portfolio Roadmap is composed of a set of
Operational Improvements
•
Each OI indicates a particular step towards achieving one or more
of the JPDO key capabilities (e.g., trajectory-based operations) and
thus achieving one or more of the JPDO national goals (e.g.,
capacity)
•
The SMAD models groups of OI’s and individual OI’s to evaluate
the performance of the NextGen
•
Not all OIs have been modeled
–
–
Some are too vague
Some cannot be addressed by our current models
8
Systems Modeling and Analysis Division
SMAD Modeling and Analysis Framework
IPTs
Operational
Improvement (OI)
Development
OI
Impact
Characterization
SMAD
Define Future
Schedule and
Conditions
Decisions
Strategy Evaluation
Direct NAS Effects
(ACES, LMINET,
ProbTFM
Strategy
Impact
(Metrics)
Boeing Airport
Capacity Constraints)
(TSAM, AvDemand)
Multi-year
Consumer, Carrier
Ramifications
(NAS Strategy
Simulator, USCAP)
Safety,
Environmental
(NIRS, INM, EDMS)
Security,
Economic Impacts
(GRA AMMS &
NACBA,
Validation
Existing Data
(e.g. ETMS Schedules)
9
Systems Modeling and Analysis Division
SMAD Modeling and Simulation Tools
•
ACES (NASA-Ames/Sensis): Agent-based simulation of individual aircraft flying one day
of NAS activity
•
LMINET (LMI): Queuing model for airports and sectors of one day of NAS activity.
•
ProbTFM Tool (Sensis): Tool for designing and evaluating probabilistic traffic flow
management in heavy weather
•
AvDemand (Sensis): Calculates future NAS demand based on FAA forecasts
•
AvAnalyst (Sensis): Analysis and visualization tool for NASA ACES simulation outputs
•
TSAM (LaRC, VaTech): Transportation Systems Analysis Model – demand generation
and NAS-wide modeling and analysis
•
NAS-Wide Environmental Impact Model (Metron, NASEIM): Detailed calculator of
noise and emissions based on individual flight trajectories from ACES
•
GRA Screening Model (GRA): For each passenger service airport, model describing
current security lanes and processing rates; may be adapted for additional lanes or
changes in processing rates
•
FAA NAS Strategy Simulator (Ventana): Multi-year, macro-level simulation of annual
system statistics of demand, NAS activity, FAA costs and revenues
•
Airport Capacity Constraints Model (Boeing): For 35 OEP airports, computes
detailed capacity as a function of runway configuration, operational procedures, and
ground infrastructure.
10
Systems Modeling and Analysis Division
Integrated Modeling and Analysis Process
SMAD has brought together best-in-class modeling and simulation
tools and expertise to support JPDO analyses
Demand
Modeling
(Sensis, LaRC, VaTech)
Airport/Airspace
Queuing Model
LMINET
(LMI)
Determine
“Feasible”
Future
Demand
Security/Econometrics
(GRA)
Physics-Based
Airport/Airspace
Analysis
Environmental
ACES
Modeling
(Sensis, ARC)
(Metron)
Wx
modeling
Individual
NGATS
Runway
Airport
Modeling Capacities
(Boeing, LMI)
Probabilistic Wx
And TFM Tool
(Sensis)
NAS Strategy Simulator
(Ventana)
NAS Economics
11
Systems Modeling and Analysis Division
Recent NAS Tradeoff Studies
•
•
•
•
Congestion modeling
Critical flights
Delay distribution
NAS performance sensitivity
12
Systems Modeling and Analysis Division
Poisson Congestion-Delay Tradeoff
f ( n) =
In
c
to reas
c a in
pa g d
ci em
ty
ra and
ti o
µ n e −µ
n!
Performance
contour
13
Systems Modeling and Analysis Division
NAS Performance Contours
Date
2006/11/16
2007/01/07
2006/11/12
2006/12/14
2006/07/04
2006/09/08
2006/11/17
Lin
NA es of
S p co
erf nst
orm an
an t
ce
Inc
r
an easi
d/o ng
r w tra
ea ffic
the
r
Traffic
Normal
Light
Light
Normal
Light
Normal
Normal
Weather
Heavy
Heavy
Moderate
Light
Moderate
Clear
Clear
2006/11/16
2007/01/07
2006/07/04
2006/09/08
2006/11/12
2006/12/14
2006/11/17
14
Systems Modeling and Analysis Division
Critical Flights
• TFM optimization via steepest gradient
– Stochastic evaluation of
forecasted congestion
• Convolve capacity and loading PDFs
Log10(Frequency)
– Rank flights by their congestion cost
– Delay / reroute flights that exceed congestion threshold
The critical flights
07/01/07
- High congestion
- Low frequency
Log10(Flight congestion cost)
15
Systems Modeling and Analysis Division
Log10(Frequency)
Delay Distribution
Log10(Frequency)
Log10(Flight congestion cost)
Log10(Flight congestion cost)
↑ #Flights delayed
↓ Delay/flight delayed
↓ Efficiency of delay
↑ Total delay
16
Systems Modeling and Analysis Division
Mean vs RMS Delay
Nov 12, 2006
RMS delay
Increased delay distribution
Mean delay
17
Systems Modeling and Analysis Division
Cost of Distributing Delay
• RMS delay can be reduced by spreading delay to
more flights
– But at the cost of increased total delay
Nov 12, 2006
$65/minute
Increased delay distribution
18
Systems Modeling and Analysis Division
Min(Delay) and
Distributed Delay Solutions
Nov 12, 2006
ETMS/ASPM
Non agile
delay distribution
No
mini n agile
mu m
Min
dela
imu
y
md
el a y
Delay distributio
n
19
Systems Modeling and Analysis Division
BACKUP
20
Systems Modeling and Analysis Division
Performance-Based TFM
Stream A
Forecasted
congestion
Stream B
Egalitarian TFM: Minimize max(delay)
Utilitarian TFM: Minimize sum(delay)
21
Systems Modeling and Analysis Division
2006/11/12
• Sunday
– Traffic: Light (42,037 IFR tracks)
– Weather: Moderate-heavy
22
Systems Modeling and Analysis Division
ProbTFM Optimization
• Is the egalitarian premise correct?
Log10(Frequency)
– We find a great variation in flight congestion cost, with a few
flights with very high costs
The critical flights
Jan 7, 07
- High congestion
- Low frequency
Log10(Flight congestion cost)
23
The policy decision needs to be informed of NAS performance relationships …
Systems Modeling and Analysis Division
Recent results
• Probabilistic CDM
24
Systems Modeling and Analysis Division
Probabilistic CDM
• Performance-based probabilistic TFM
– Premise: Flight plans and traffic schedule are a rich solution with
many constraints and preferences built-in
• Should minimize deviation from traffic schedule
– Try to minimize control effort for a given NAS performance target
Log10(Frequency)
• Give operators visibility into flight costs and the tools; let them solve the problem
The critical flights
Jan 7, 07
- High congestion
- Low frequency
Log10(Flight congestion cost)
25
Systems Modeling and Analysis Division
Prob CDM
Traffic
schedule
National Airspace System
Service provider tactical TFM
AOC Strategic TFM
26