Title
SMART mobility via prediction, optimization and personalization
Author
Atasoy, B. (TU Delft Transport Engineering and Logistics) 
Azevedo, Carlos Lima (Technical University of Denmark)
Akkinepally, Arun Prakash (Massachusetts Institute of Technology)
Seshadri, Ravi (Singapore-MIT Alliance)
Zhao, Fang (Singapore-MIT Alliance)
Abou-Zeid, Maya (American University of Beirut)
Ben-Akiva, Moshe E. (Massachusetts Institute of Technology)
Contributor
Antoniou, Constantinos (editor)
Chaniotakis, Emmanouil (editor)
Efthymiou, Dimitrios (editor)
Date
2020
Abstract
In this chapter, we present a methodological approach for Smart Mobility that integrates three key features: prediction, optimization, and personalization. They are integrated in such a way that when a travel menu is offered, predicted conditions are considered in the attributes of alternatives and optimized system-level policies are maintained. Similarly, user-level estimations and updates are used by prediction and optimization methods at the system-level in order to represent the population with most up-to-date behavioral estimates. Furthermore, a simulation-based evaluation methodology enables to validate the performance of prediction, optimization, and personalization before Smart Mobility is implemented in real-life. Two case studies are presented based on the proposed methodologies together with platforms that facilitate their application. Potential benefits of the proposed methodologies are evaluated which can be classified into user-level and system-level benefits. User-level benefits include consumer surplus, waiting times, etc., and system-level is concerned with congestion, throughput, system-wide travel time, etc. As there is normally a tradeoff between the individual decision-making and system-wide decision-making, Smart Mobility bridges them together with appropriate methodologies on each end. For example, for our Flexible Mobility on Demand case study, we observe 10%–20% reduction in volume-to-capacity ratio as a system-level benefit. Moreover, we see that the tradeoff between consumer surplus and operator profit can be managed with an appropriate objective function.
Subject
Agent-based simulation
Dynamic traffic assignment
Inter- and intraconsumer heterogeneity
Online calibration
Optimization
Personalization
Prediction
Smart mobility
To reference this document use:
http://resolver.tudelft.nl/uuid:f7885e5e-3fab-4311-af1e-47ab987498bb
DOI
https://doi.org/10.1016/B978-0-12-815018-4.00012-7
Publisher
Elsevier
Embargo date
2020-06-06
ISBN
978-0-12-815019-1
Source
Demand for Emerging Transportation Systems: Modeling Adoption, Satisfaction, and Mobility Patterns
Bibliographical note
Green Open Access added to TU Delft Institutional Repository 'You share, we take care!' - Taverne project https://www.openaccess.nl/en/you-share-we-take-care Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public.
Part of collection
Institutional Repository
Document type
book chapter
Rights
© 2020 B. Atasoy, Carlos Lima Azevedo, Arun Prakash Akkinepally, Ravi Seshadri, Fang Zhao, Maya Abou-Zeid, Moshe E. Ben-Akiva