The University of Toronto ITE Student Chapter welcomed special guest speaker Professor Moshe Ben-Akiva of MIT on March 19, 2021 to present “Tri-POP: An online platform for smart mobility with prediction, optimization and personalization.”
UTTRI Director, Professor Eric Miller, introduced Professor Ben-Akiva, saying:
“Even the most cursory summary of his accomplishments and contributions to the field of transportation are overwhelming. Professor Ben-Akiva has been a global leader and a driving force in advancing transportation demand modelling and systems analysis research and practice for his long and storied career dating all the way back to his seminal PhD thesis work.” – Professor Eric Miller
Ben-Akiva shared that he “almost came to University of Toronto 50 years ago” at the invitation of Professor Richard Soberman.
He opened by saying that his talk was about operations of smart mobility. He divided his presentation into three sections:
- Smart mobility solutions that his research group has worked on;
- Tri-POP analytics, talking in detail about the Tri-POP platform; and
- Tri-POP applications, summarizing three applications that his research group has been engaged in.
Ben-Akiva began with Flexible Mobility On Demand (FMOD) as an example of a smart mobility solution that his group worked on with Fujitsu about ten years ago. The goal was to improve public transportation. The project focused on paratransit services, specifically, private taxi (or solo ride hailing), shared taxi (e.g. UberPool), and mini-bus (on-demand). Bringing together these three alternatives, his research group developed an algorithm that, given the fleet of vehicles, produces alternatives that fall into these three categories, depending on availability of the fleet.
Using a smartphone app, users request a trip – providing origin, destination and departure time – and the app returns pick-up time, travel time and price for up to three options.
Ben-Akiva showed the type of results obtained from simulation experiments using two different objective functions: one to maximize profit, and one to maximize consumer surplus (benefits) with a fixed fleet size. The operator, on one hand, is interested in maximizing profit. The traveller, on the other hand, is interested in shorter wait and travel times. Simulation using profit maximization would generate 20% more profit. Optimizing consumer surplus would generate less revenue but more consumer surplus.
“The outcome depends on what you are optimizing,” he said. Is the point of view the operator, and profit? Or is the point of view the traveller, and maximizing consumer surplus?
Ben-Akiva noted that “waiting time actually decreases when you maximize profit, because the operator will tend to provide the door-to-door services that are more profitable.”
He went on to explain Tri-POP in detail and discuss several other examples: Tri-POP for managed lanes; Tripod, which offers personalized incentives for sustainability through a trip planner app; and Freight on Demand (FOD). He spoke briefly about the naming of Tri-POP:
“Where does the name Tri-POP come from? The POP comes from Prediction, Optimization, Personalization. Or, you can think that this is TriP OPtimization. But to be honest, I was based, before the pandemic, in Singapore, and I travelled to Korea, where I heard a lot about K-pop – so I was inspired by that!” – Professor Moshe Ben-Akiva
In conclusion, Professor Ben-Akiva summarized Tri-POP’s benefits:
“Tri-POP feeds intelligent analytics into the operations of smart mobility to gain improvements in service, in revenue, in efficient use of resources, and in user satisfaction.” – Professor Moshe Ben-Akiva
UT-ITE President Mostafa (Yaser) Kouchakzadeh moderated Q & A following the seminar.
- Watch the presentation video recording on YouTube “Tri-POP: An online platform for smart mobility with prediction, optimization and personalization,” Professor Moshe Ben-Akiva, March 19, 2021.
- View Professor Moshe Ben-Akiva’s presentation file PDF.
Abstract
Tri-POP is an online platform for operations of smart mobility solutions. In its core it combines online analytics for prediction, optimization, and personalization. It uses a bi-level optimization of system and user levels. The system optimization problem is solved periodically (e.g., every 5 minutes) to determine the optimal policy (e.g., pricing, fleet rebalancing, incentive allocation) for attaining system-level objectives (e.g., travel time, welfare, revenue).
In our applications the online prediction and the system level optimization are performed by DynaMIT, a simulation-based Dynamic Traffic Assignment (DTA) system that combines historical traffic data and real-time surveillance information through online calibration. The user optimization problem is triggered at every user request to determine a customized menu of options based on individual-level preferences. Online Bayesian inference is employed to update users’ preferences as users make choices. Machine learning algorithms infer the user choices using smartphone sensors. In our applications the user level algorithms are embedded in the Future Mobility Sensing (FMS) platform. Tri-POP is illustrated through applications to flexible mobility on demand, sustainable travel incentives, personalized tolling of managed lanes, and deliveries on demand.
Simulation experiments demonstrate the potential benefits of Tri-POP to regulators, operators and users.
About the speaker
Moshe Ben-Akiva is the Edmund K. Turner Professor of Civil and Environmental Engineering at the Massachusetts Institute of Technology (MIT), Director of the MIT Intelligent Transportation Systems Lab, and Principal Investigator at the Singapore-MIT Alliance for Research and Technology.
He holds a PhD degree in Transportation Systems from MIT and was awarded honorary degrees from the University of the Aegean, the Université Lumiére Lyon, the KTH Royal Institute of Technology, and the University of Antwerp. His awards include the Robert Herman Lifetime Achievement Award in Transportation Science from the Institute for Operations Research and the Management Sciences, the Lifetime Achievement Award of the International Association for Travel Behavior Research, the Jules Dupuit prize from the World Conference on Transport Research Society, and the Institute of Electrical and Electronics Engineers ITS Society Outstanding Application Award for DynaMIT, a system for dynamic network management.
Ben-Akiva has co-authored two books, including the textbook Discrete Choice Analysis, published by MIT Press, and nearly 400 papers in refereed journals or refereed conferences. He has worked as a consultant in industries such as transportation, energy, telecommunications, financial services and marketing for a number of private and public organizations, including Hague Consulting Group, RAND Europe, and Cambridge Systematics, where he was previously a Senior Principal and member of the Board of Directors. He also was an advisor to Memetrics and ChoiceStream, provided litigation support to Analysis Group and Brattle Group and is the Chief Scientific Advisor to Mobile Market Monitor. He was recently a member of the Future Interstate Highway System Committee of the National Academies of Sciences, Engineering, and Medicine.
Presented by University of Toronto ITE Student Chapter, UT-ITE.