Research Overview

I am broadly interested in the design of service systems and the accompanying policies and mechanisms (both operational, e.g., routing, matching, and staffing policies, and economic, e.g., pricing, cost/revenue sharing, and incentive mechanisms) that involve complex interactions among different strategic entities whose objectives do not necessarily align with that of the system manager. An emerging focus of my research is to better understand trade-offs between competing economic objectives, and how managing these trade-offs fundamentally constrains the underlying operational decisions and vice versa. This focus is largely motivated by the global transformation of urban mobility into a user-centric, service-oriented playing field, from a traditionally vehicle-centric, infrastructure-oriented framework. Technology advances that enable Mobility-as-a-Service (MaaS) and the sharing economy, transform the underlying problem space and expand the design space along new dimensions. Consequently, my research involves adapting traditional modelling frameworks and methodologies accordingly, and analyzing them to identify key insights that can guide policy and decision-making in large, real-world systems such as transportation.

My approach is to first characterize the limitations of purely systemic design choices (i.e., excluding external monetary dependencies), and then identify the key trade-offs when also considering (limited) monetary dependencies (e.g., performance-based rewards/penalties, government subsidies, etc.) on the design. I combine tools from multiple subjects such as optimization, probability, queueing theory, decision theory, game theory and mechanism design for obtaining analytical results, as well as numerical methods and, more recently, data-driven simulation for validation in real-world environments. My recent work can be classified under the following topics:

  • Service Operations Management

    Service Systems with Strategic Agents

    Studying the impact of strategic customer/server behavior on optimal service system design, which involves the choice of system configuration (pooled with one central queue or dedicated with multiple parallel queues), the routing policy (e.g., Random routing or Fastest Server First), the matching policy (which customers should be served together), and the staffing policy (how many servers to hire).       More

  • Pricing & Revenue Management

    Sustainable Market Design for Modern Urban Mobility

    Identifying and addressing the key challenges and trade-offs in designing a welfare-maximizing urban mobility market mechanism, while simultaneously encouraging participation from utility-maximizing commuters, non-profit public transit agencies, and profit-sensitive on-demand mobility services. The ability of the market to potentially combine services from multiple mobility service providers to offer efficient multi-modal solutions to commuters, thereby injecting complementarity into otherwise purely substitute services, can be exploited to possibly overcome some well-known fundamental limitations.                 More

    Sequential Individual Rationality and Fairness

    Utilitarian approach to modeling QoS by introducing sequential notions of IR and fairness that favorably regulate fluctuations in the expected utility of customers during service at decision epochs to optimize QoS-aware pricing (by a commercial service provider) or cost sharing (among customers who independently share service) in dynamic shared service systems such as ridesharing that seek to incentivize the choice of shared rather than exclusive service.          More

    Characterizing Distribution Rules for Cost Sharing Games

    How the collective cost incurred or revenue generated by the actions of self-interested agents should be shared among them, to guarantee stable and efficient outcomes in a broad class of resource allocation games.            More