Monte Carlo AI for Air Traffic. Kagan Tumer MCAI 2013 March 20, 2013

Size: px
Start display at page:

Download "Monte Carlo AI for Air Traffic. Kagan Tumer MCAI 2013 March 20, 2013"

Transcription

1 Monte Carlo AI for Air Traffic Kagan Tumer MCAI 2013 March 20, 2013

2 Air Traffic Conges:on

3 Traffic: Interesting Problem? Problem inherently large and distributed Many cars Many intersections Many planes Good decisions change with time Daily scale Yearly scale Solution needs to be adaptive Sensor lights aren t very bright Solution needs to scale: One intersection A hundred intersections Huge amount of uncertainty Weather Events

4 Language Agent: an algorithm that will sense, decide, act on the environment Multiagent systems: Many agents interact Interactions create unpredictable learning environment Air traffic is a great example of a large, stochastic multiagent systems

5 Intelligent Agents in Traffic Highway traffic Drivers Lanes City Traffic Lights Cars (not drivers) Air Traffic Gates Routes Aircraft Role of air traffic controllers

6 Air Traffic Flow Management Current Situation 40,000+ flights operate in the US airspace in one day Delays caused by weather and airport conditions: 1,682,700 hours of delay (2007) 740,000,000 gallons of fuel wasted (2007) Estimated cost impact: over $41 billion (2007) Moving forward Threefold increase in air traffic Increased heterogeneity of aircraft Need Algorithmic solution Infrastructure will not change significantly

7 Current Air Traffic Management Air Traffic decisions made at four levels: 1. Airspace Management (6 hours to 1 year) Game Plan Centralized 2. National Flow (2-8 hours) Centralized 3. Regional Flow (20 min-2 hours) Hierarchical 4. Separation Assurance (2-30 minutes) Air traffic controllers

8 Current Air Traffic Management Air Traffic decisions made at four levels: 1. Airspace Management (6 hours to 1 year) Game Plan Centralized 2. National Flow (2-8 hours) Centralized 3. Regional Flow (20 min-2 hours) Hierarchical 4. Separation Assurance (2-30 minutes) Air traffic controllers

9 Multiagent Learning for Air Traffic? Advantages: Large distributed problem Naturally decentralized Human senses are overwhelmed by data Challenges: Humans have to remain in the loop Agent approach needs to be transparent Allow humans to take over Help humans don t replace them

10 Snapshot of the airspace

11 First steps What are we measuring? System performance? (reward/objective/utility/evaluation) How are we measuring it? System snapshots (state) System movie What about system dynamics? Simulators

12 What are we after? How do we know if we succeed? Define a system level objective Minimize congestion What about delays?

13 System Objective Function Minimize congestion C(z) = C s (z) s S C s (z) = t (k s,t c s ) 2 I ks,t >c s Minimize delays B(z) = B (z) a a A B a (z) = (t a τ a ) I ta >τ a

14 System Objective Function Full state vector Lateness-congestion tradeoff coefficient Lateness Term Congestion Term

15 Simulation: FACET

16 FEATS: Fast Event-based Air Traffic Simulator Simulation of generic or US airspace Developed at Oregon State University Can simulate up to 28,000 flights per second Designed for Monte Carlo simulations and learning Learning may occur: After a complete simulation At discrete times during the simulation (use the LEARN event) By triggering some other event Simulation: FEATS

17 We need 4 more things Learning Multiagent Approach

18 Agent-Based Air Traffic Management 1. Identify agents 2. Identify actions 3. Derive agent objective functions 4. Select agent learning algorithm

19 Approach: Identify Agents Agents as aircraft? agents Little data to train agents Actions conflict with pilots Agents as routes? Not well defined agents Actions of routes? Agents as fix locations? Number of agents vary with need All flight plans contain at least one agent fix. Agents have simple actions: set metering restrictions Agents can be active or inactive (e.g., live around congestion).

20 Approach: Identify Agents Agent Fix min MIT min MIT min MIT Agents as fix locations? Number of agents vary with need All flight plans contain at least one agent fix. Agents have simple actions: set metering restrictions Agents can be active or inactive (e.g., live around congestion).

21 Agent-Based Air Traffic Management 1. Identify agents Fixes 2. Identify actions Miles in Trail Ground holds Reroutes 3. Derive agent objective functions Difference objective 4. Select agent learning algorithm Simple reinforcement learning

22 Basic Algorithm An agent keeps table of Values for each ac:on: V(a) Policy: With probability epsilon choose random ac:on Otherwise choose ac:on with highest value Agent takes an ac:on and receives a reward R Value update: V(a) (1 α) V(a) + α R

23 Difference Reward Look at difference between system reward, and system reward with agent taking constant action c i D i ( z) = G( z) G( z + c ) i i System Reward System Reward Without i s influence D is hard to compute: D requires n + 1 runs of FACET for every learning episode G requires 1 run Solution: Estimate difference reward

24 Difference Reward Look at difference between system reward, and system reward with agent taking constant action c i D i ( z) = G( z) G( z + c ) i i Key theoretical result: G(z i + c i ) z i = 0 g i (z) z i = G(z) z i

25 Difference Reward Look at difference between system reward, and system reward with agent taking constant action c i D i ( z) = G( z) G( z + c ) i i Key theoretical result: D and G are aligned: What s good for me is good for the system

26 Experiments & Algorithms Experiments: Two Artificially Created Congestions One heavy congestion One light congestion 300 aircraft Algorithms: Monte Carlo Estimation Q-learning agents using: G - system reward D - difference reward (local) Dest1 - first estimate of difference reward Dest2 - second estimate of difference reward

27 Results: 20 Agents

28 Penalty Tradeoff Delay Only Congestion Only

29 Scaling Maximum System Utility Achieved D D est G Monte Carlo Number of Agents

30 More advanced results What if there are humans in the loop? Suggestion agents What if you can t compute the objective functions? Model system objectives

31 Agent Decision Making Input Agent Ac:on Reward Input Agent Intended Ac:on Noise Actual Ac:on Actual Reward Reward Received by Agent Noise

32 Human in the Loop: Suggestion Agents Input Agent Intended Ac:on Noise Agent Ac:on Filter Actual Ac:on Actual Reward Reward Received by Agent Noise

33 Suggestion Agents Maximum System Reward Achieved Agent Solution Minimizing Congestion and Lateness Agents Give Suggestions (using D) Agents Give Suggestions (using G) Human Models Minimizing Congestion Only Number of Steps

34 Suggestion Agents Final Maximum System Reward Achieved Agents Give Suggestions (using D) Agents Give Suggestions (using G) Agents Only Human Models Only Weight of Agents Suggestions

35 What if you can t compute the objectives? Use a function approximation Tabular linear function Non-linear approximator (like a neural network)

36 Modeling Objectives

37 Modeling Objectives

38 Summary of Air Traffic Project Lessons Learned: Key of successful application is in selecting: Agents Obvious choice is not always the best Agent actions Modify system parameters Agent rewards Good choice has direct impact on performance Don t rock the boat too much: Humans have to remain in the loop Agent approach needs to be transparent Allow humans to take over

39 Acknowledgements and Publica:ons Collabora:ve work spanning over 7 years NASA Ames Research Center, UARC Adrian Agogino Oregon State University ScoQ Proper, A:l Iscen, Carrie Rebhuhn, Will Curran, Zach Welch Select papers: A Mul:agent Approach to Managing Air Traffic Flow. Journal of Autonomous Agents and Mul2agent Systems, 24:1-25, 2012 Learning Indirect Ac:ons in Complex Domains: Ac:on Sugges:ons for Air Traffic Control. Advances in Complex Systems, 12: , Regula:ng Air Traffic Flow with Coupled Agents. In Proceedings of the Seventh Interna2onal Joint Conference on Autonomous Agents and Mul2agent Systems. Estoril, Portugal, May Distributed Agent- Based Air Traffic Flow Management. In Proceedings of the Sixth Interna2onal Joint Conference on Autonomous Agents and Mul2agent Systems, pages Honolulu, HI, May NSF support: Grants CNS and CNS- II- EN

40 Ques:ons? Kagan Tumer hqp://web.engr.oregonstate.edu/~ktumer/ AADI: Autonomous Agents and Distributed Intelligence Lab

Learning to Trick Cost-Based Planners into Cooperative Behavior

Learning to Trick Cost-Based Planners into Cooperative Behavior Learning to Trick Cost-Based Planners into Cooperative Behavior Carrie Rebhuhn, Ryan Skeele, Jen Jen Chung, Geoffrey A. Hollinger and Kagan Tumer Abstract In this paper we consider the problem of routing

More information

Overcoming Communication Restrictions in Collectives

Overcoming Communication Restrictions in Collectives Overcoming Communication Restrictions in Collectives Kagan Tumer Computational Sciences Division NASA Ames Research Center Mailstop 269-4 Moffett Field, CA 94035 E-mail: ktumer@mail.arc.nasa.gov Adrian

More information

Rule based approach. to traffic management. Copyright LibRT 2016

Rule based approach. to traffic management. Copyright LibRT 2016 Rule based approach to traffic management 1 It's all about the road user's experience who is driving a smarter car with more technology. 2 SUMMERY NETWORK: POLICY DECISION Flow control points: {A n } ROAD

More information

Predicting gas usage as a function of driving behavior

Predicting gas usage as a function of driving behavior Predicting gas usage as a function of driving behavior Saurabh Suryavanshi, Manikanta Kotaru Abstract The driving behavior, road, and traffic conditions greatly affect the gasoline consumption of an automobile.

More information

Improving Traffic Flow Optimization and Demand Management

Improving Traffic Flow Optimization and Demand Management Improving Traffic Flow Optimization and Demand Management - Impact of uncertainty on decision making - Modeling economic effects - Rerouting as an operational solution - Mathematical modeling for optimal

More information

Commission hearing on the preparation for RP3

Commission hearing on the preparation for RP3 Commission hearing on the preparation for RP3 14th December 2016 Brussels Discussion document Introduction The performance scheme is at the heart of Single European Sky (SES) policy and is a major driver

More information

Thus, there are two points to keep in mind when analyzing risk:

Thus, there are two points to keep in mind when analyzing risk: One-Minute Spotlight WHAT IS RISK? Uncertainty about a situation can often indicate risk, which is the possibility of loss, damage, or any other undesirable event. Most people desire low risk, which would

More information

Design of Aircraft Trajectories based on Trade-offs between Emission Sources

Design of Aircraft Trajectories based on Trade-offs between Emission Sources Design of Aircraft Trajectories based on Trade-offs between Emission Sources Banavar Sridhar and Neil Chen NASA Ames Research Center Moffett Field, CA, USA Hok Ng University of California, Santa Cruz,

More information

Evaluation of Congestion Pricing for Management Highway in Seattle

Evaluation of Congestion Pricing for Management Highway in Seattle Evaluation of Congestion Pricing for Management Highway in Seattle Aman J. Aman, Diana C. Nuñez, Mohammed Hamideddin, Pankaj Laungani, Wail Abbo Abstract Congestion is one of the most important and complex

More information

Collaborative Decision Making in Aviation

Collaborative Decision Making in Aviation the way we do it Collaborative Decision Making in Aviation In an uncertain world there is reasonable certainty in stating that air travel will increase in the future. How airlines and airports will work

More information

ACTAM: Cooperative Multi-Agent System Architecture for Urban Traffic Signal Control

ACTAM: Cooperative Multi-Agent System Architecture for Urban Traffic Signal Control ACTAM: Cooperative Multi-Agent System Architecture for Urban Traffic Signal Control SIB Sunil Gyawali Isaac Vargas & Benjamin Bertrand Outline Introduction Objective of our Seminar Multi-Agent System in

More information

Support of Traffic Management Coordinators in an Airport Air Traffic Control Tower Using the Surface Management System

Support of Traffic Management Coordinators in an Airport Air Traffic Control Tower Using the Surface Management System From: HCI-02 Proceedings. Copyright 2002, AAAI (www.aaai.org). All rights reserved. Support of Traffic Management Coordinators in an Airport Air Traffic Control Tower Using the Surface Management System

More information

mycopter Enabling Technologies for Personal Aerial Transportation Systems

mycopter Enabling Technologies for Personal Aerial Transportation Systems mycopter Enabling Technologies for Personal Aerial Transportation Systems Prof. Dr. Heinrich H. Bülthoff Max Planck Institute for Biological Cybernetics Tübingen, Germany Project funded by the European

More information

Telematics System dedicated for Provisioning of On-Line Information about Predicted Occupancy of Highway Parking Areas

Telematics System dedicated for Provisioning of On-Line Information about Predicted Occupancy of Highway Parking Areas Telematics System dedicated for Provisioning of On-Line Information about Predicted Occupancy of Highway Parking Areas Tomas ZELINKA, Michal JERABEK, Martin SROTYR, Zdenek LOKAJ, Petr KUMPOST Faculty of

More information

Risk Analysis Overview

Risk Analysis Overview Risk Analysis Overview What Is Risk? Uncertainty about a situation can often indicate risk, which is the possibility of loss, damage, or any other undesirable event. Most people desire low risk, which

More information

Analysis of Multi-Sector Planner Concepts in U.S. Airspace

Analysis of Multi-Sector Planner Concepts in U.S. Airspace Analysis of Multi-Sector Planner Concepts in U.S. Airspace Kevin Corker, Paul Lee, Tom Prevot, Eromi Guneratne, Lynne Martin, Nancy Smith, Savvy Verma, Jeffrey Homola, and Joey Mercer San Jose State University

More information

AEM 5495 Spring Design, Build, Model, Simulate, Test and Fly Small Uninhabited Aerial Vehicles (UAVs)

AEM 5495 Spring Design, Build, Model, Simulate, Test and Fly Small Uninhabited Aerial Vehicles (UAVs) AEM 5495 Spring 2011 Design, Build, Model, Simulate, Test and Fly Small Uninhabited Aerial Vehicles (UAVs) Gary J. Balas balas@umn.edu Monday-Wednesday 3:35-4:50 PM 211 Akerman Hall UAV Course Syllabus

More information

PROBABILISTIC CONGESTION MANAGEMENT

PROBABILISTIC CONGESTION MANAGEMENT PROBABILISTIC CONGESTION MANAGEMENT Craig Wanke, Lixia Song, Stephen Zobell, Daniel Greenbaum, and Sandeep Mulgund The MITRE Corporation, cwanke@mitre.org, McLean, VA Abstract A new approach to airspace

More information

WebShipCost - Quantifying Risk in Intermodal Transportation

WebShipCost - Quantifying Risk in Intermodal Transportation WebShipCost - Quantifying Risk in Intermodal Transportation Zhe Li, Heather Nachtmann, and Manuel D. Rossetti Department of Industrial Engineering University of Arkansas Fayetteville, AR 72701, USA Abstract

More information

Project Overview: SMALL AIR TRANSPORT ROADMAP

Project Overview: SMALL AIR TRANSPORT ROADMAP http://sat-rdmp.eu/ Project Overview: In the SAT-Rdmp (Support Action) a Vision and Research Agenda on the development of a Small Aircraft Transportation system (SAT system) have been developed. The major

More information

JPDO Systems Modeling and Analysis

JPDO Systems Modeling and Analysis JPDO Systems Modeling and Analysis Dr. George Hunter Presented on behalf of Yuri Gawdiak JPDO Systems Modeling and Analysis Division (SMAD) September 5, 2007 JPDO Organizational Changes Yuri Gawdiak has

More information

Potential Operational Benefits of Multi-layer Point Merge System on Dense TMA Operation

Potential Operational Benefits of Multi-layer Point Merge System on Dense TMA Operation Potential Operational Benefits of Multi-layer Point Merge System on Dense TMA Operation Hybrid Arrival Trajectory Optimization Applied to Beijing Capital International Airport Man LIANG, Daniel DELAHAYE

More information

OJT INSTRUCTOR. Learning outcomes. Why study this course? Aim. ICAO Code 212

OJT INSTRUCTOR. Learning outcomes. Why study this course? Aim. ICAO Code 212 OJT INSTRUCTOR ICAO Code 212 ATC Operational Training 10 days Air Traffic Controllers The younger generation expects to know how they are achieving and they want to engage and take responsibility for their

More information

Traffic Study for Deerfoot Trail, Calgary

Traffic Study for Deerfoot Trail, Calgary Transportation Engineering Department, Schulich School of Engineering Traffic Study for Deerfoot Trail, Calgary Capacity Drop, Hysteresis, MFD and METANET Model for Deerfoot Trail Karan Arora UCID: 10141627

More information

TRANSPORTATION FACTS. OUR CUSTOMERS: Travel Patterns

TRANSPORTATION FACTS. OUR CUSTOMERS: Travel Patterns OUR CUSTOMERS: Travel Patterns TRANSPORTATION FACTS The more than 6.5 million people in the Commonwealth have many reasons to travel and several ways to get to their destinations. A recent statewide survey

More information

Capacity Planning and Assessment Additional considerations

Capacity Planning and Assessment Additional considerations Capacity Planning and Assessment Additional considerations Air Traffic Services System Capacity Seminar/Workshop Nairobi, Kenya, 8 10 June 2016 Raffaele Russo EUROCONTROL Operations Planning Summary Winter

More information

Collaborative Decision Making (CDM) Update

Collaborative Decision Making (CDM) Update Collaborative Decision Making (CDM) Update 1 FAA CDM Office CDM Domestic combined with International Operations, includes International Data Sharing One Manager Two Full Time Team Leads, Two Part-Time

More information

SIEMENS CONCERT: City of Seattle

SIEMENS CONCERT: City of Seattle SIEMENS CONCERT: City of Seattle Use of ATMS to Enhance and Extend Signal Operations Abstract: Today s signal systems are part of a complex environment that includes varied and disparate data sources and

More information

Potential Benefits of Operations

Potential Benefits of Operations Destination Green ICAO Symposium on Aviation and Climate Change, Destination Green, 14 16 May 2013 Potential Benefits of Operations John-Paul B. Clarke Associate Professor Daniel Guggenheim School of Aerospace

More information

providing businesses with the ability to learn more about their operating environment, and identify and act with the potential to create new value.

providing businesses with the ability to learn more about their operating environment, and identify and act with the potential to create new value. More and more of the world s activity is being expressed digitally by billions of interconnected devices Smarter Cities Smarter Energy Smarter Vehicles Smarter Factories Smarter Homes Smarter Health Smarter

More information

REAL TIME ASSESSMENT OF DRINKING WATER SYSTEMS USING A DYNAMIC BAYESIAN NETWORK

REAL TIME ASSESSMENT OF DRINKING WATER SYSTEMS USING A DYNAMIC BAYESIAN NETWORK REAL TIME ASSESSMENT OF DRINKING WATER SYSTEMS USING A DYNAMIC BAYESIAN NETWORK W. J. Dawsey 1, B. S. Minsker 2 and E. Amir 3 1 Graduate Research Asst., Dept of Civil & Environmental Engineering, University

More information

ATC BASIC. Learning outcomes. Why study this course? Aim. ICAO Code 051

ATC BASIC. Learning outcomes. Why study this course? Aim. ICAO Code 051 ATC BASIC ICAO Code 051 6 weeks Trainee ATCs (Maximum 12 per course) This course forms the prerequisite for all other Air Traffic Control Courses. trainee will have demonstrated competency with regards

More information

Origin-Destination Trips and Skims Matrices

Origin-Destination Trips and Skims Matrices Origin-Destination Trips and Skims Matrices presented by César A. Segovia, AICP Senior Transportation Planner AECOM September 17, 2015 Today s Webinar Content Matrix Estimation Data sources Florida Application

More information

How to Best Automate Intersection Management

How to Best Automate Intersection Management How to Best Automate Intersection Management Aashiq Parker Department of Computer Science University of Cape Town Cape Town, South Africa PRKAAS003@myuct.ac.za Geoff Nitschke Department of Computer Science

More information

NextGen ATM Concept of Operations: ASAS-Reliant

NextGen ATM Concept of Operations: ASAS-Reliant NextGen ATM Concept of Operations: ASAS-Reliant See http://www.jpdo.gov for the latest JPDO info Doug Arbuckle, Rose Ashford NextGen Joint Planning & Development Office 18-Sep-2007 Joint 5th ASAS TN2 Workshop

More information

GUIDELINES FOR SELECTING TRAVEL FORECASTING METHODS AND TECHNIQUES

GUIDELINES FOR SELECTING TRAVEL FORECASTING METHODS AND TECHNIQUES GUIDELINES FOR SELECTING TRAVEL FORECASTING METHODS AND TECHNIQUES Maren Outwater and Jeff Doyle, RSG, 55 Railroad Row, White River Junction, VT 05001 maren.outwater@rsginc.com, jeff.doyle@rsginc.com ABSTRACT

More information

Review of system identification and control algorithms used for smart motorways applications

Review of system identification and control algorithms used for smart motorways applications Review of system identification and control algorithms used for smart motorways applications Olivier Haas Mobility and Transport Research Centre, Coventry University, CV1 2TL, UK Acknowledgement: Abiodun

More information

Improvement on the Acceptance of a Conflict Resolution System by Air Traffic Controllers. R. Flicker, Technical University Berlin

Improvement on the Acceptance of a Conflict Resolution System by Air Traffic Controllers. R. Flicker, Technical University Berlin Improvement on the Acceptance of a Conflict Resolution System by Air Traffic Controllers R. Flicker, Technical University Berlin Outline 1. Background 1. Multi-Sector-Planning 2. Development of Conflict

More information

ISE480 Sequencing and Scheduling

ISE480 Sequencing and Scheduling ISE480 Sequencing and Scheduling INTRODUCTION ISE480 Sequencing and Scheduling 2012 2013 Spring term What is Scheduling About? Planning (deciding what to do) and scheduling (setting an order and time for

More information

Dynamic Lane Merge Systems

Dynamic Lane Merge Systems Dynamic Lane Merge Systems Acknowledgements Maryland State Highway Administration University of Maryland, College Park Michigan Department of Transportation Wayne State University Lane Merging Issues Merging

More information

Factored Customer Models for Agent-based Smart Grid Simulation

Factored Customer Models for Agent-based Smart Grid Simulation Factored Customer Models for Agent-based Smart Grid Simulation Prashant Reddy Manuela Veloso Machine Learning Department Computer Science Department School of Computer Science Carnegie Mellon University

More information

Automatic Development Tools. Abstract

Automatic Development Tools.  Abstract Automatic Development Tools in Software Engineering Courses Janusz Zalewski School of Electrical Engineering & Computer Science University of Central Florida Orlando, FL 32816-2450, USA jza@ece.engr.ucf.edu

More information

Variable Speed Warnings on DMS

Variable Speed Warnings on DMS Variable Speed Warnings on DMS Evolving Tools for Better Safety and Service Jeanne Olubogun, P.E. District Traffic Engineer - St. Louis Region Jeanne.Olubogun@modot.mo.gov 314-275-1536 2014 Traffic and

More information

FLEXIBLE APPOINTMENT BASED SYSTEM WITH ADAPTIVE RESPONSE TO TRAFFIC AND PROCESSING DELAYS

FLEXIBLE APPOINTMENT BASED SYSTEM WITH ADAPTIVE RESPONSE TO TRAFFIC AND PROCESSING DELAYS FLEXIBLE APPOINTMENT BASED SYSTEM WITH ADAPTIVE RESPONSE TO TRAFFIC AND PROCESSING DELAYS Amrinder Arora, DSc NTELX Research and Development, 1945 Old Gallows Rd, Suite 700, McLean VA 22182, USA +1 703

More information

The Effectiveness of Risk Management Within the DoD

The Effectiveness of Risk Management Within the DoD The Effectiveness of Risk Management Within the DoD Funded by the Naval Postgraduate School, Dr. Keith Snider Donald McKeon, PhD Sterling Heights, MI 586-276-2172 don.mckeon@dau.mil Slide 1 of 35 Report

More information

RELATIONSHIP OF MAXIMUM MANAGEABLE AIR TRAFFIC CONTROL COMPLEXITY AND SECTOR CAPACITY

RELATIONSHIP OF MAXIMUM MANAGEABLE AIR TRAFFIC CONTROL COMPLEXITY AND SECTOR CAPACITY 26 TH INTERNATIONAL CONGRESS OF THE AERONAUTICAL SCIENCES RELATIONSHIP OF MAXIMUM MANAGEABLE AIR TRAFFIC CONTROL COMPLEXITY AND SECTOR CAPACITY Parimal Kopardekar*, Jessica Rhodes**, Albert Schwartz**,

More information

How the IoT and Data Analytics Will Drive the Future of the Supply Chain

How the IoT and Data Analytics Will Drive the Future of the Supply Chain How the IoT and Data Analytics Will Drive the Future of the Supply Chain Dan Romary VP, Data Analytics Presented by: 2018 MHI Copyright claimed for audiovisual works and sound recordings of seminar sessions.

More information

Safer travel, faster arrival siemens.com/mobility

Safer travel, faster arrival siemens.com/mobility Sitraffic Conduct+ Highway Management System Safer travel, faster arrival siemens.com/mobility Less congestion. Fewer accidents. Higher capacity. 2 Wherever overhead gantries display situation-tailored

More information

Airport Collaborative Decision Making Enhancing Airport Efficiency

Airport Collaborative Decision Making Enhancing Airport Efficiency Airport Collaborative Decision Making Enhancing Airport Efficiency David Gamper Director, Safety & Technical ACI World SWIFT Conference, Banff 18 September 2012 1 Airports are facing challenges Capacity

More information

Intelligent continuous improvement, when BPM meets AI. Miguel Valdés Faura CEO and co-founder

Intelligent continuous improvement, when BPM meets AI. Miguel Valdés Faura CEO and co-founder Intelligent continuous improvement, when BPM meets AI Miguel Valdés Faura CEO and co-founder BPM IS NOT DEAD. But we should admit though, BPM has been a bit unsexy lately. And exactly, what is your job

More information

Artificial Intelligence in Automotive Production

Artificial Intelligence in Automotive Production Chair Information Management in Mechanical Engineering Artificial Intelligence in Automotive Production 3. Fachkonferenz: Roboter in der Automobilindustrie Stuttgart, Germany, 29 th November 2017 Robotic

More information

Earthquake Rescue Mission Modeling Based on Multi- Agent

Earthquake Rescue Mission Modeling Based on Multi- Agent TELKOMNIKA Indonesian Journal of Electrical Engineering Vol.12, No.4, April 2014, pp. 2995 ~ 3000 DOI: http://dx.doi.org/10.11591/telkomnika.v12i4.4999 2995 Earthquake Rescue Mission Modeling Based on

More information

Global Supply Chain Planning under Demand and Freight Rate Uncertainty

Global Supply Chain Planning under Demand and Freight Rate Uncertainty Global Supply Chain Planning under Demand and Freight Rate Uncertainty Fengqi You Ignacio E. Grossmann Nov. 13, 2007 Sponsored by The Dow Chemical Company (John Wassick) Page 1 Introduction Motivation

More information

Usability Evaluation of the Spot and Runway Departure Advisor (SARDA) Concept in a Dallas/Fort Worth Airport Tower Simulation

Usability Evaluation of the Spot and Runway Departure Advisor (SARDA) Concept in a Dallas/Fort Worth Airport Tower Simulation Usability Evaluation of the Spot and Runway Departure Advisor (SARDA) Concept in a Dallas/Fort Worth Airport Tower Simulation Miwa Hayashi, Ty Hoang, Yoon Jung NASA Ames Research Center Gautam Gupta, Waqar

More information

REAL TIME SIMULATION OF INTEGRATION OF UAV'S INTO AIRSPACE

REAL TIME SIMULATION OF INTEGRATION OF UAV'S INTO AIRSPACE 26 TH INTERNATIONAL CONGRESS OF THE AERONAUTICAL SCIENCES REAL TIME SIMULATION OF INTEGRATION OF UAV'S INTO AIRSPACE Dirk-Roger Schmitt, Sven Kaltenhäuser, Bernd Keck Deutsches Zentrum für Luft- und Raumfahrt

More information

Probabilistic Methodology For Life Prediction of Aircraft Turbine Rotors

Probabilistic Methodology For Life Prediction of Aircraft Turbine Rotors Probabilistic Methodology For Life Prediction of Aircraft Turbine Rotors Abstract M. P. Enright 1, L. Huyse 2, R. C. McClung 3, and H.R. Millwater 4 The presence of rare metallurgical or manufacturing

More information

Management Information Systems, Sixth Edition. Chapter 10: Decision Support and Expert Systems

Management Information Systems, Sixth Edition. Chapter 10: Decision Support and Expert Systems Management Information Systems, Sixth Edition Chapter 10: Decision Support and Expert Systems Objectives List and explain the phases in decision making Articulate the difference between structured and

More information

2.0 DESCRIPTION OF PROPOSED ACTION AND ALTERNATIVES

2.0 DESCRIPTION OF PROPOSED ACTION AND ALTERNATIVES 2.0 DESCRIPTION OF PROPOSED ACTION AND ALTERNATIVES 2.0 DESCRIPTION OF PROPOSED ACTION AND ALTERNATIVES 2.1 PROPOSED ACTIONS, FORT GREELY, ALASKA The additional GMD VOC activities analyzed in this supplemental

More information

Simulation as Support for PBL Contract Design

Simulation as Support for PBL Contract Design Simulation as Support for PBL Contract Design Olle Wijk Systecon Rehnsgatan 20 104 32 Stockholm +46 8 459 0757 olle.wijk@systecon.se Patric Andersson Systecon Rehnsgatan 20 104 32 Stockholm +46 8 459 0766

More information

CANSO Global Approach to Safety. Sam Espig NATS

CANSO Global Approach to Safety. Sam Espig NATS CANSO Global Approach to Safety Performance Measurement Sam Espig NATS Agenda Measures and targets CANSO Safety Strategy ICAO Aviation Safety Index Slide 2 Why set targets? Provide common direction Motivate

More information

The AIRPORT CDM. Bahrain October 2015

The AIRPORT CDM. Bahrain October 2015 The AIRPORT CDM Bahrain 11-13 October 2015 Airport CDM a definition Airport CDM is a proven concept which aims at improving predictability, reducing delays, optimally utilizing the available capacities

More information

Dr. Cynthia West Executive Director Office of Sustainability & Climate Change

Dr. Cynthia West Executive Director Office of Sustainability & Climate Change Dr. Cynthia West Executive Director Office of Sustainability & Climate Change 193 million acres 500 million acres 57,000 miles of trails 4,300 campgrounds $13 billion Federal Forests Private Forests

More information

2018 ncode User Group Meeting

2018 ncode User Group Meeting 2018 ncode User Group Meeting February 28 March 1, 2018 Novi, MI USA Analysis of Connected Vehicle Data Dr. Andrew Halfpenny Director of Technology ncode Products Contents 3 1. What are Connected Vehicles

More information

Global Manufacturing Industry Landscape

Global Manufacturing Industry Landscape Global Manufacturing Industry Landscape Tim Hanley, Deloitte Global Leader, Consumer & Industrial Products Industry group 2017 Manufacturing Forum. 17 October 2017. Perdue University What we will cover

More information

Terms of Reference. Purchase of aerial surveillance service for the EU external land borders

Terms of Reference. Purchase of aerial surveillance service for the EU external land borders Annex II Terms of Reference Purchase of aerial surveillance service for the EU external land borders Ref. No: Frontex/ Purchase of aerial border surveillance service aircraft equipped with multi-intelligence

More information

A Thesis Presented to The Academic Faculty. Anuj P. Shah

A Thesis Presented to The Academic Faculty. Anuj P. Shah Analysis of Transformations to Socio-Technical Systems Using Agent-based Modeling and Simulation A Thesis Presented to The Academic Faculty by Anuj P. Shah In Partial Fulfillment of the Requirements for

More information

US Coast Guard Modeling/Simulation Efforts: Foreign Fishing Vessel (FFV) Incursions

US Coast Guard Modeling/Simulation Efforts: Foreign Fishing Vessel (FFV) Incursions US Coast Guard Modeling/Simulation Efforts: Foreign Fishing Vessel (FFV) Incursions LCDR D. Blair Sweigart LCDR E. Taquechel USCG Atlantic Area, Operations Analysis Division 30 June 2015 Objective Problem:

More information

AGENT-BASED SIMULATIONS USING HUMAN PERFORMANCE MODELS FOR NATIONAL AIRSPACE SYSTEM RISK ASSESSMENT. Annual Performance Report For the Period

AGENT-BASED SIMULATIONS USING HUMAN PERFORMANCE MODELS FOR NATIONAL AIRSPACE SYSTEM RISK ASSESSMENT. Annual Performance Report For the Period "76- AGENT-BASED SIMULATIONS USING HUMAN PERFORMANCE MODELS FOR NATIONAL AIRSPACE SYSTEM RISK ASSESSMENT Annual Performance Report For the Period submitted to NASA Ames Research Center Moffet Field, CA

More information

SESAR development and deployment

SESAR development and deployment SESAR development and deployment Global Challenges to Improve Air Navigation Performance Asilomar Conference Grounds, Pacific Grove, CA 13 February 2015 Rolf TUCHHARDT European Commission, DG MOVE SESAR

More information

SMART HIGHWAYS & TUNNELS

SMART HIGHWAYS & TUNNELS A LEADING LAYER SWARCO specializes in the development of smart solutions in transport and infrastructure, and is a leading player in the market for intelligent transportation systems. Our solutions are

More information

The Nottingham eprints service makes this work by researchers of the University of Nottingham available open access under the following conditions.

The Nottingham eprints service makes this work by researchers of the University of Nottingham available open access under the following conditions. Kilsby, P. and Remenyte-Prescott, Rasa and Andrews, John D. (2016) Modelling asset management of railway overhead line equipment. In: 9th IMA International Conference on Modelling in Industrial Maintenance

More information

NAS-Wide Performance: Impact of Select Uncertainty Factors and Implications for Experimental Design

NAS-Wide Performance: Impact of Select Uncertainty Factors and Implications for Experimental Design Approved for Public Release: 12-0370. Distribution Unlimited. NAS-Wide Performance: Impact of Select Uncertainty Factors and Implications for Experimental Design Gareth O. Coville, MITRE Corporation Billy

More information

Federal Aviation Administration FPAW Weather Integration Into NextGen

Federal Aviation Administration FPAW Weather Integration Into NextGen FPAW Weather Integration Into NextGen Date: October 12, 2011 Presented By: Shirley Burgess Chief System Engineer En Route and Oceanic Services PMO Update ATO Program Management Organization (PMO) Congressional

More information

Towards Collaborative Mobility: Elastic Transport Infrastructure & Research Challenges

Towards Collaborative Mobility: Elastic Transport Infrastructure & Research Challenges Towards Collaborative Mobility: Elastic Transport Infrastructure & Research Challenges 20 th ITS World Congress Angelos Amditis, ICCS Tokyo, October 15th, 2013 This project is co-funded by the European

More information

OPTIMIZING RAMP METERING STRATEGIES

OPTIMIZING RAMP METERING STRATEGIES OPTIMIZING RAMP METERING STRATEGIES Presented by Kouros Mohammadian, Ph.D. Saurav Chakrabarti. ITS Midwest Annual Meeting Chicago, Illinois February 7, 2006 Background Ramp control is the application of

More information

Driving Compliance with Functional Safety Standards for Software-Based Automotive Components

Driving Compliance with Functional Safety Standards for Software-Based Automotive Components Driving Compliance with Functional Safety Standards for Software-Based Automotive Components EXECUTIVE SUMMARY T oday s automobile is a technology hub on wheels, with connected systems and embedded software

More information

1. For s, a, initialize Q ( s,

1. For s, a, initialize Q ( s, Proceedings of the 2006 Winter Simulation Conference L. F. Perrone, F. P. Wieland, J. Liu, B. G. Lawson, D. M. Nicol, and R. M. Fujimoto, eds. A REINFORCEMENT LEARNING ALGORITHM TO MINIMIZE THE MEAN TARDINESS

More information

Building a Safety Case for Automated Mobility: Smart Cities and Autonomous Mobility Getting There Safely

Building a Safety Case for Automated Mobility: Smart Cities and Autonomous Mobility Getting There Safely Building a Safety Case for Automated Mobility: Smart Cities and Autonomous Mobility Getting There Safely Building a Safety Case for Automated Mobility: Smart Cities and Autonomous Mobility Getting There

More information

Stephen Ween and Jane Knipe. Drones The flight path from toy shop to research tool

Stephen Ween and Jane Knipe. Drones The flight path from toy shop to research tool Stephen Ween and Jane Knipe Drones The flight path from toy shop to research tool Outline We will outline: What a drone is Legislative requirements and responsibilities Summary of requirements and general

More information

Copyright 2014 Oracle and/or its affiliates. All rights reserved.

Copyright 2014 Oracle and/or its affiliates. All rights reserved. Oracle Buys TOA Technologies Adds Leading Field Service SaaS to Oracle Service Cloud and Oracle ERP Cloud Solutions to Deliver Effective and Timely Home- and Facility-based Customer Service September 17,

More information

Transportation Modeling Tools. Rainer Dronzek November 3, 2006

Transportation Modeling Tools. Rainer Dronzek November 3, 2006 Transportation Modeling Tools Rainer Dronzek November 3, 2006 Presenter Info BSEE, Bradley University Institute of Industrial Engineers simulation Ask the Expert Simulation Solutions Conference chair and

More information

SESAR: a European initiative for an enhanced ATM system

SESAR: a European initiative for an enhanced ATM system domaine Recherche & SESAR: a European initiative for an enhanced ATM system General overview and spotlight on the SESAR operational concept Patrick DUJARDIN R&D Manager DSNA/DTI Toulouse - France Tokyo,

More information

ICAT MIT. Increase Milestone Prediction Accuracy in. Potential Approaches to. NAS Movements. Georg Theis, Francis Carr, Professor JP Clarke

ICAT MIT. Increase Milestone Prediction Accuracy in. Potential Approaches to. NAS Movements. Georg Theis, Francis Carr, Professor JP Clarke M I T I n t e r n a t i o n a l C e n t e r f o r A i r T r a n s p o r t a t i o n M I T I n t e r n a t i o n a l C e n t e r f o r A i r T r a n s p o r t a t i o n Potential Approaches to Increase

More information

Air Traffic Management Capacity-Driven Operational Concept Through 2015

Air Traffic Management Capacity-Driven Operational Concept Through 2015 2 nd USA/EUROPE AIR TRAFFIC MANAGEMENT R&D SEMINAR Orlando,1 st - 4 th December 1998 Air Management Capacity-Driven Operational Concept Through 2015 Aslaug Haraldsdottir Robert W. Schwab Monica S. Alcabin

More information

Advanced Information Systems Big Data Study for Earth Science

Advanced Information Systems Big Data Study for Earth Science Advanced Information Systems Big Study for Earth Science Daniel Crichton, NASA Jet Propulsion Laboratory Michael Little, NASA Headquarters October 29, 2015 Background NASA has historically focused on systematic

More information

Intelligent Agents in Mobile Vehicular Ad-Hoc Networks: Leveraging Trust Modeling Based on Direct Experience with Incentives for Honesty

Intelligent Agents in Mobile Vehicular Ad-Hoc Networks: Leveraging Trust Modeling Based on Direct Experience with Incentives for Honesty 2010 IEEE/WIC/ACM International Conference on Web Intelligence and Intelligent Agent Technology Intelligent Agents in Mobile Vehicular Ad-Hoc Networks: Leveraging Trust Modeling Based on Direct Experience

More information

DISTRIBUTED ARTIFICIAL INTELLIGENCE

DISTRIBUTED ARTIFICIAL INTELLIGENCE DISTRIBUTED ARTIFICIAL INTELLIGENCE LECTURE 3: PROBLEM MODELING INTRODUCTION There are different approaches for modeling of agent-based systems. The model that has gained most attention ti is modeling

More information

Decision Support and Business Intelligence Systems

Decision Support and Business Intelligence Systems Decision Support and Business Intelligence Systems (9 th Ed., Prentice Hall) Chapter 4: Modeling and Analysis Learning Objectives Understand the basic concepts of management support system (MSS) modeling

More information

AIM APPLICATION PLATFORM INTELLIGENT MOBILITY

AIM APPLICATION PLATFORM INTELLIGENT MOBILITY AIM APPLICATION PLATFORM INTELLIGENT MOBILITY Sascha Bauer, Eike Bretschneider, Jürgen Rataj, Tilo Schuman DLR - Institute of Transport Systems Rutherfordstr. 2, 12489 Berlin Germany +49-30-670-55-409,

More information

A neural reinforcement learning approach to learn local dispatching policies in production scheduling

A neural reinforcement learning approach to learn local dispatching policies in production scheduling A neural reinforcement learning approach to learn local dispatching policies in production scheduling Simone Riedmiller Institut fur Werkzeugmaschinen und Betriebstechnik (wbk) University of Karlsruhe

More information

Operations in the 21st Century DOT Meeting Customers Needs and Expectations

Operations in the 21st Century DOT Meeting Customers Needs and Expectations Operations in the 21st Century DOT Meeting Customers Needs and Expectations 1 Purpose of this Meeting: Share Thoughts & Discuss Challenges brought about by the changing transportation environment and public

More information

Chapter 2 Performance Measures

Chapter 2 Performance Measures Transportation Decision-making Principles of Project Evaluation and Programming Chapter 2 Performance Measures K. C. Sinha and S. Labi Purdue University School of Civil Engineering 1 Performance Measures

More information

Industry 4.0 and Automotive Digital Transformation. Jim Bydalek May 9 th, 2017

Industry 4.0 and Automotive Digital Transformation. Jim Bydalek May 9 th, 2017 Industry 4.0 and Automotive Digital Transformation Jim Bydalek May 9 th, 2017 This automotive industry is rapidly evolving Jacques Aschenbroich Valeo Volkmar Denner, Bosch Elmar Degenhart, Continental

More information

Self-adaptive Distributed Software Systems

Self-adaptive Distributed Software Systems Self-adaptive Distributed Software Systems INF 5360 spring 2015 lecturer: Amir Taherkordi INF5360/9360 spring 2015: overview self-adaptive software systems 1 Overview Ø Preliminary definitions Ø Motivation

More information

Metaheuristics and Cognitive Models for Autonomous Robot Navigation

Metaheuristics and Cognitive Models for Autonomous Robot Navigation Metaheuristics and Cognitive Models for Autonomous Robot Navigation Raj Korpan Department of Computer Science The Graduate Center, CUNY Second Exam Presentation April 25, 2017 1 / 31 Autonomous robot navigation

More information

Deviations Prediction in Timetables Based on AVL Data

Deviations Prediction in Timetables Based on AVL Data V. Kůrková et al. (Eds.): ITAT 2014 with selected papers from Znalosti 2014, CEUR Workshop Proceedings Vol. 1214, pp. 54 60 http://ceur-ws.org/vol-1214, Series ISSN 1613-0073, c 2014 Z. Jiráček, V. Martínek,

More information

Testing Air Traffic Controllers Cognitive Complexity Limits. Design Proposal Fall Author: Howard Kleinwaks

Testing Air Traffic Controllers Cognitive Complexity Limits. Design Proposal Fall Author: Howard Kleinwaks Testing Air Traffic Controllers Cognitive Complexity Limits Design Proposal 16.621 Fall 2001 Author: Howard Kleinwaks Advisor: R. John Hansman Partner: Blake Gottlieb December 11, 2001 1 Table of Contents

More information

BUS Public Transportation System Fuel Efficiency Patterns

BUS Public Transportation System Fuel Efficiency Patterns BUS Public Transportation System Fuel Efficiency Patterns José de Almeida, and João C Ferreira Abstract This paper provides an overview of a set of OLAP and Data Mining based feedback tools for management

More information

Market Based Allocation of Power in Smart Grid

Market Based Allocation of Power in Smart Grid 11 American Control Conference on O'Farrell Street, San Francisco, CA, USA June 29 - July 1, 11 Market Based Allocation of Power in Smart Grid Baisravan HomChaudhuri and Manish Kumar, Member, ASME Abstract

More information

Machina Research White Paper for ABO DATA. Data aware platforms deliver a differentiated service in M2M, IoT and Big Data

Machina Research White Paper for ABO DATA. Data aware platforms deliver a differentiated service in M2M, IoT and Big Data Machina Research White Paper for ABO DATA Data aware platforms deliver a differentiated service in M2M, IoT and Big Data December 2013 Connections (billion) Introduction More and more businesses are making

More information