skip to main content
US FlagAn official website of the United States government
dot gov icon
Official websites use .gov
A .gov website belongs to an official government organization in the United States.
https lock icon
Secure .gov websites use HTTPS
A lock ( lock ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites.

Explore Research Products in the PAR
It may take a few hours for recently added research products to appear in PAR search results.


Title: Contextual Bandit-Based Sequential Transit Route Design under Demand Uncertainty
While public transit network design has a wide literature, the study of line planning and route generation under uncertainty is not so well covered. Such uncertainty is present in planning for emerging transit technologies or operating models in which demand data is largely unavailable to make predictions on. In such circumstances, this paper proposes a sequential route generation process in which an operator periodically expands the route set and receives ridership feedback. Using this sensor loop, a reinforcement learning-based route generation methodology is proposed to support line planning for emerging technologies. The method makes use of contextual bandit problems to explore different routes to invest in while optimizing the operating cost or demand served. Two experiments are conducted. They (1) prove that the algorithm is better than random choice; and (2) show good performance with a gap of 3.7% relative to a heuristic solution to an oracle policy.  more » « less
Award ID(s):
1652735
PAR ID:
10213709
Author(s) / Creator(s):
;
Date Published:
Journal Name:
Transportation Research Record: Journal of the Transportation Research Board
Volume:
2674
Issue:
5
ISSN:
0361-1981
Page Range / eLocation ID:
613 to 625
Format(s):
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. ; ; ; ; ; (, 2017 IEEE 56th Annual Conference on Decision and Control (CDC))
    Urban public transit planning is crucial in reducing traffic congestion and enabling green transportation. However, there is no systematic way to integrate passengers' personal preferences in planning public transit routes and schedules so as to achieve high occupancy rates and efficiency gain of ride-sharing. In this paper, we take the first step tp exact passengers' preferences in planning from history public transit data. We propose a data-driven method to construct a Markov decision process model that characterizes the process of passengers making sequential public transit choices, in bus routes, subway lines, and transfer stops/stations. Using the model, we integrate softmax policy iteration into maximum entropy inverse reinforcement learning to infer the passenger's reward function from observed trajectory data. The inferred reward function will enable an urban planner to predict passengers' route planning decisions given some proposed transit plans, for example, opening a new bus route or subway line. Finally, we demonstrate the correctness and accuracy of our modeling and inference methods in a large-scale (three months) passenger-level public transit trajectory data from Shenzhen, China. Our method contributes to smart transportation design and human-centric urban planning. 
    more » « less
  2. ; (, Travel Behaviour and Society)
    With the advent of new mobility modes and technologies, we have seen meaningful changes in travel behavior. One such new mobility mode is on-demand transit. The Metropolitan Atlanta Rapid Transit Authority deployed its own on-demand transit system, dubbed MARTA Reach, in March of 2022. This paper provides an evaluation of the characteristics of two groups of people related to MARTA Reach: those who were interested in it and used it and those who were interested in it but did not use it. In addition, this paper explores the factors that influence membership in each of those two groups using a binary logit model, revealing the underlying characteristics that are linked with the decision to use or not use the service given prior interest. The findings show that simply providing more service has the strongest effect on adoption. Among 561 survey respondents, 426 expressed that the service area for MARTA Reach was too limited for their needs. Modeling results support this finding, in addition to the following strong predictors of on-demand transit adoption: 1) being a frequent transit user, 2) being satisfied with the current state of fixed-route transit service, 3) being part of a low-income household, 4) living within an on-demand transit service area, and 5) being younger. Understanding these group characteristics and underlying factors can help guide future efforts to provide on-demand transit service, such as by targeting the market segments that share features with the underlying factors that are shown herein to be linked with on-demand transit adoption. 
    more » « less
  3. ; ; (, Networks)
    Abstract Local delivery networks expect drivers to make deliveries to and/or pickups from customers using the shortest routes in order to minimize costs, delivery time, and environmental impact. However, in real‐world applications, it is often the case that not all customers are known when planning the initial delivery route. Instead, additional customers become known while the driver is making deliveries or pickups. Before serving the new demand requests, the vehicle will return to the depot for restocking. In other words, there exists a precedence relation in the delivery route to visit the depot before delivering new orders. The uncertainty in new customer locations can lead to expensive rerouting of the tour, as drivers revisit previous neighborhoods to serve the new customers. We address this issue by constructing the delivery route with the knowledge that additional customers will appear, using historical demand patterns to guide our predictions for the uncertainty. We model this network delivery problem as a precedence‐constrained asymmetric traveling salesman problem using mixed‐integer optimization. Experimental results show that the proposed robust optimization approach provides an effective delivery route under the uncertainty of customer demands. 
    more » « less
  4. ; ; ; (, INFORMS Transportation and Logistics Society Second Triennial Conference)
    Current free and subscription-based trip planners have heavily focused on providing available transit options to improve the first and last-mile connectivity to the destination. However, those trip planners may not truly be multimodal to vulnerable road users (VRU)s since those selected side walk routes may not be accessible or feasible for people with disability. Depending on the level of availability of digital twin of travelers behaviors and sidewalk inventory, providing the personalized suggestion about the sidewalk with route features coupled with transit service reliability could be useful and happier transit riders may boost public transit demand/funding and reduce rush hour congestion. In this paper, the adaptive trip planner considers the real-time impact of environment changes on pedestrian route choice preferences (e.g., fatigue, weather conditions, unexpected construction, road congestion) and tolerance level in response to transit service uncertainty. Side walk inventory is integrated in directed hypergraph on the General Transit Feed Specification to specify traveler utilities as weights on the hyperedge. A realistic assessment of the effect of the user-defined preferences on a traveler’s path choice is presented for a section of the Boston transit network, with schedule data from the Massachusetts Bay Transportation Authority. Different maximum utility values are presented as a function of varying traveler’s risk-tolerance levels. In response to unprecedented climate change, poverty, and inflation, this new trip planner can be adopted by state agencies to boost their existing public transit demand without extra efforts 
    more » « less
  5. ; ; (, Transportation Science)
    The rise of on-demand mobility technologies over the past decade has sparked interest in the integration of traditional transit and on-demand systems. One of the main reasons behind this is the potential to address a fundamental trade-off in transit: the ridership versus coverage dilemma. However, unlike purely fixed systems or purely on-demand systems, integrated systems are not well understood; their planning and operational problems are significantly more challenging, and their broader implications are the source of a heated debate. Motivated by this debate, we introduce the dynamicity gap, a general concept that quantifies the attainable benefit of allowing (but not requiring) dynamic components in the response strategy to a multistage optimization problem. Although computing the dynamicity gap exactly may be intractable, we develop an analytical framework with which to approximate it as a function of problem input parameters. The framework allows us to certify the value of dynamism (i.e., a dynamicity gap greater than one) for certain combinations of problem input parameters. We showcase our approach with two sets of computational experiments, from which we gain both qualitative and quantitative insights about the settings in which the integration of transit and on-demand systems may certifiably be a worthwhile investment. Funding: This work was partially supported by the National Science Foundation [Grants DMS-1839346 and CNS-1952011]. Part of this research was performed while the authors were visiting the Institute for Pure and Applied Mathematics, which is supported by the National Science Foundation [Grant DMS-1925919]. Supplemental Material: The online appendix is available at https://doi.org/10.1287/trsc.2022.1193 . 
    more » « less
  • Citation Formats
  • MLA
  • APA
  • Chicago
  • BibTeX
  • Save / Share this Record
  • Send to Email