Autonomous and Connected Transportation Systems: Modeling, Control, and Deployment

ITSC 2019 Workshop

Date and time: 09:00-17:00, Sunday, October 27, 2019

Organizers: Mauro Salazar, Ramon Iglesias, Stephen Zoepf, Marco Pavone

The workshop will start at 9 AM.


Public debate about the future of mobility and transportation is increasingly informed by predictions about the impact of Autonomous Vehicles (AVs). As AVs are approaching market-readiness, it becomes more critical that we answer questions about them:

  • How can we design profitable and sustainable mobility systems that leverage AVs?
  • What will these new forms of mobility and transportation mean for society?
  • How can we ensure that such technologies benefit all members of society, improving equity rather than undermining it?
This workshop will gather experts from transportation, operations research, robotics, and urban planning in order to:
  1. identify challenges and opportunities for the future of transportation that are triggered by the advent of AVs,
  2. identify modeling and control methodologies to address them,
  3. share insights from early deployments and turn such insights into an actionable research roadmap.


Speaker Time Title Abstract
Mauro Salazar 09:00 - 09:30 Autonomous Mobility-on-Demand Systems for Future Urban Mobility

In this talk I will discuss the operational and societal aspects of autonomous mobility-on-demand (AMoD) systems, a rapidly developing mode of transportation wherein mobility is provided on demand by robotic, self-driving vehicles. Specifically, I will discuss AMoD systems along three dimensions: (1) modeling, namely mathematical frameworks capable of capturing the salient dynamic and stochastic features of customer demand, (2) control, that is coordination algorithms for the vehicles aimed at throughput maximization, and (3) societal, entailing system-level studies characterizing the interaction between AMoD and other infrastructures, such as the electric power and public transit networks. I will conclude the talk by presenting a number of directions for future research.

Krishna Selvam 09:30 - 10:00 Ride Sharing Marketplace: Designing for Efficiency

Designing an efficient marketplace for ride sharing is a challenging task, since several independent systems need to respond in real time without getting in each other’s way. While the nature of supply in an autonomous world might be very different from what it is today, there are still several learnings and fundamental principles that are likely to be relevant in the future. This talk will describe how Lyft approaches marketplace design with respect to dynamic pricing, dispatch and driver positioning and how we plan for an autonomous future.

--- 10:00 - 10:30 ---

Coffee Break

Francesco Ciari 10:30 - 11:15 Planning Shared Automated Vehicle Fleets: Specific Modeling Requirements and Concepts to Address Them

Autonomous vehicles, in recent years, quickly became one of the most popular research topics in transportation. Several scientists proposed a vision in which shared autonomous vehicles fleets would be used instead of privately owned cars and would serve the current transportation demand with an extremely reduced number of vehicles. However, the implications that such a dramatic change of paradigm would have on transportation planning itself, were neglected almost entirely so far. Multi-modal simulation tools have been enhanced to accommodate autonomous vehicles travel but, if automation will actually help shared mobility becoming prevalent, more changes to current planning approaches will be needed. We might need to look at different sets of problems, and further enhance the models we use in order to solve them. This talk, points at some simple but important aspects of individuals’ travel behavior that imply inherent limitations of existing approaches. It also proposes ideas to adapt existing tools, or develop new ones, more suitable for a world in which autonomous shared vehicles would be the backbone of the transportation system.

Raphael Stern 11:15 - 12:00 Controlling Mixed Human and Autonomous Traffic

This work is motivated by the possibility of a small number of autonomous vehicles (AVs) or partially autonomous vehicles that may soon be present on our roadways, and the impacts they will have on traffic flow. This automation may take the form of fully autonomous vehicles without human intervention (SAE Level 5) or, as is already the case in many modern vehicles, may take the form of driver assist features such as adaptive cruise control (ACC) or other SAE Level 1 features. Regardless of the extent of automation, changing the vehicle dynamics of a small number of vehicles in the bulk traffic flow may have substantial implications on the underlying traffic dynamics and may influence traffic stability and the development of emergent phenomena such as phantom traffic jams.

In this talk, I present some recent experimental work conducted to understand how AVs may be able to influence traffic flow as well as a series of experiments to understand the car following behavior of commercially available ACC vehicles. The two-vehicle car following data is collected, and models are calibrated to describe the vehicle-level ACC car following behavior. These models are then analyzed for string stability. Of the seven vehicle models tested, all are found to be string unstable. To validate the models for multi-vehicle simulations, a platoon test with seven ACC following vehicles of the same make and model is also conducted, and stability analysis of the calibrated models is conducted.

--- 12:00 - 14:00 ---

Lunch Break

Michael Levin 14:00 - 14:30 Maximum-stability Dispatch Policy for Shared Autonomous Vehicles

Shared autonomous vehicle (SAVs) are predicted to be a major usage model for automated vehicles. By providing driverless point-to-point service, SAV ridership costs could be similar to personal vehicle ownership. Consequently, recent studies have performed extensive simulations to investigate the effects on travel service, fleet size, and congestion. Unfortunately, a major limitation is in the dispatch strategy for matching vehicles to passengers. Vehicle routing problems are NP-hard, and the SAV problem is complicated by uncertainty in future demand. Consequently, studies have relied on heuristics but their level of optimality is unknown, which limits the generality of the simulation results. This paper investigates an analytical max-pressure policy which aims to maximize passenger throughput under stochastic demand. Although we make no claims about the optimality of waiting times, this policy is proven to maintain bounded queues of waiting passengers if at all possible. The stability region results can be used to determine the minimum fleet size needed for given average demand.

Javier Alonso-Mora 14:30 - 15:00 Predictive Routing and Multi-objective Fleet Sizing for Shared Mobility-on-demand

We move towards an era of smart cities, where autonomous cars will provide on-demand transportation while making our roads safer. In this talk I will give an overview of our work in dynamic vehicle routing for large-scale ride-sharing. Firstly, for fleet management, I will describe an anytime optimal method that is capable of online assignment of large numbers of requests to vehicles and routing them accordingly. The proposed framework includes rebalancing of idle vehicles and has been extended towards predictive routing with a model of future demand. Secondly, I will discuss a multi-objective optimisation to trade-off quality of service vs operation cost.

--- 15:00 - 15:30 ---

Coffee Break

Emilio Frazzoli 15:30 - 16:00 Autonomous Mobility-on-Demand: What is Known and What is Not Known

In past years, many of the most urgent questions related to autonomous mobility-on-demand were purely technical and vehicle centered: how safe are these vehicles, how comfortable are they, what scenarios can they handle? As public deployments worldwide are becoming more common, capabilities and limitations are becoming clearer. On the other hand, much less confirmed knowledge is available on the operation and impact of large-scale autonomous mobility-on-demand systems with thousands of vehicles. What are the key parameters and constraints on operation? How many seats should the robotic taxis have? What will the influence on congestion be? Answering these questions is an exercise that requires not only technical know-how but also the consideration of societal and economic aspects. While much is still in the dark, research of most recent years has answered some of the open questions. This talk will provide insights into these answers of what we already know about autonomous mobility-on-demand. Moreover, it will also highlight some of the aspects which until today remain unanswered.

--- 16:00 - 16:30 Discussion on Future Research Directions

Update: Michal Čáp and Salomon Wollenstein-Betech had to cancel last-minute due to unforeseen circumstances.


Javier Alonso-Mora

Javier Alonso-Mora is an Assistant Professor at the Cognitive Robotics department of the Delft University of Technology. Previously he was a Postdoctoral Associate at the Computer Science and Artificial Intelligence Lab (CSAIL) of MIT. He received his Ph.D. degree in robotics from ETH Zurich, in partnership with Disney Research Zurich. His main research interest is in navigation, motion planning and control of autonomous mobile robots and vehicle fleets. He is the recipient of an VENI award from the Netherlands Organisation for Scientific Research (2017) and a Best paper award in multi-robot systems at the IEEE ICRA (2019).

Michal Čáp

Michal Čáp is a technology lead at ISEE AI and a research associate at Artificial Intelligence Center, Czech Technical University in Prague. He received Bc. degree in Information Technology from Brno University of Technology, the Czech Republic and MSc degree in Agent Technology from Utrecht University, the Netherlands. In 2017, he obtained a PhD degree in Artificial Intelligence from CTU in Prague, Czech Republic. Between years 2015 and 2016, Michal Čáp held the position of Fulbright-funded visiting researcher at Massachusetts Institute of Technology, USA. In 2018, Michal was a postdoctoral fellow at the Department of Cognitive Robotics, TU Delft, the Netherlands. His research focuses on multi-robot motion planning and autonomous transportation systems.

Francesco Ciari

Dr. Francesco Ciari is Assistant Professor at Polytehcnique de Montréal. He obtained his PhD in transportation planning at ETH Zurich and his research mainly focuses on the modeling and assessment of innovative transportation systems, in particular shared mobility. He has over ten years of experience in agent-based transportation modeling, thanks to his involvement in the development of the simulation software MATSim. He is also co-founder and chief innovation officer of GeoTwin, a Paris based start-up, which offers consulting and develops software for the planning of innovative mobility solutions.

Emilio Frazzoli

Emilio Frazzoli is a professor of Dynamic Systems and Control at ETH Zurich, and Chief Scientist of Aptiv’s Autonomous Mobility unit. His main research interest are in robotics, autonomous systems, and intelligent mobility. In acknowledgement of his seminal work in these fields, Emilio has received numerous awards, including the the 2015 IEEE George S. Axelby Award and the 2017 IEEE Kiyo Tomiyasu Award, and has been named an IEEE Fellow in 2019. Emilio has published more than 200 papers in the fields of robotics, autonomous vehicles, and drones. A former full professor at MIT, he directed the research group that first demonstrated an autonomous mobility (“robotaxi”) service to the public, and performed the first analysis of the social and economic impact of such a service, based on real transportation data. In 2013 he founded nuTonomy with Karl Iagnemma, and served as its Chief Technology Officer until its acquisition by Aptiv in 2017.

Michael Levin

Michael W. Levin is an Assistant Professor in the Department of Civil, Environmental, and Geo- Engineering at the University of Minnesota. He received a B.S. degree in Computer Science and a Ph.D. degree in Civil Engineering from The University of Texas at Austin in 2013 and 2017, respectively. He is a recipient of the Dwight D. Eisenhower Fellowship from the Federal Highway Administration and the 2016 Milton Pikarsky Award from the Council of University Transportation Centers. Dr. Levin is a member of the Network Modeling Committee (ADB30) of the Transportation Research Board. His research focuses on traffic flow and network modeling of connected autonomous vehicles and intelligent transportation systems.

Mauro Salazar

Mauro Salazar is a Postdoctoral Scholar at the Autonomous Systems Lab in the Department of Aeronautics and Astronautics at Stanford University. He received the Ph.D. degree in Mechanical Engineering from ETH Zurich in 2019. Mauro’s research is at the interface of control theory and optimization, and is aimed at the development of a comprehensive set of tools for the design, the deployment and the operation of future mobility systems. Specifically, his area of expertise includes optimal control theory, hybrid electric vehicles, and autonomous mobility-on-demand. Mauro received the Outstanding Bachelor Award and the Excellence Scholarship and Opportunity Award from ETH Zurich. His Master thesis was recognized with the ETH Medal. He was awarded the Best Student Paper award at the 2018 Intelligent Transportation Systems Conference.

Krishna Selvam

Krishna Kumar Selvam has been at Lyft for 3 years, working as a Research Scientist to build and improve Lyft’s dispatch systems on nearly every ride type. He has also worked on proof of concept dispatch and positioning algorithms for a partially autonomous fleet. Krishna holds Master’s degrees in Transportation and Highway Engineering from MIT and Infrastructure Civil Engineering from IIT Madras.

Raphael Stern

Dr. Raphael Stern is an assistant professor in the department of Civil, Environmental, and Geo- Engineering at the University of Minnesota. Prior to joining UMN, Dr. Stern spent time as a postdoctoral researcher in the Department of Informatics at the Technical University of Munich. Dr. Stern received a bachelor of science degree (2013), master of science degree (2015), and Ph.D. (2018) all in Civil Engineering from the University of Illinois at Urbana-Champaign. Dr. Stern was a visiting researcher at the Institute for Pure and Applied Mathematics at UCLA, a visiting researcher at the Institute for Software Integrated Systems at Vanderbilt University, and a recipient of the Dwight David Eisenhower Graduate Fellowship from the Federal Highway Administration. Dr. Stern’s research interests are in the area of traffic control and estimation with autonomous vehicles in the flow.

Salomón Wollenstein-Betech

Salomón Wollenstein-Betech is a Ph.D. Candidate in the Division of Systems Engineering at Boston University. He is part of the CODES lab under the supervision of Prof. Christos Cassandras and the NOC lab under the supervision of Prof. Ioannis Paschalidis and is the recipient of the Dean’s Fellowship awarded by the university. His research focuses on data-driven assessment, optimization and control of Intelligent Transportation Systems with a focus on traffic congestion. Among other projects, Salomón has also collaborated with the MIT-IBM Watson AI Lab working on the interpretability of black-box controllers with an application on smart traffic light controllers.