Abhay Sobhanan
Portrait of Abhay Sobhanan

Abhay Sobhanan

I am an Assistant Professor in the Decision Sciences area and a Young Faculty Research Chair at the Indian Institute of Management Bangalore. I develop operations research and machine learning methods for complex optimization problems in transportation, warehousing and supply chain management.

I lead the CORBIT Lab. I completed my PhD in Industrial Engineering at the University of South Florida in July 2025, advised by Hadi Gard and Changhyun Kwon. My dissertation developed methods for large-scale vehicle routing problems and took on practical concerns such as allocating workload fairly. I also hold a BS–MS dual degree in Mathematics from the National Institute of Technology Agartala (2019).

News

  1. Sep 2026

    Received the IIMB Challenge Grant.

  2. Apr 2026

    Received the ANRF Prime Minister Early Career Research Grant.

  3. Jan 2026

    Paper on workload equity in crowdshipping accepted for publication in Production and Operations Management.

  4. Aug 2025

    Joined the Indian Institute of Management Bangalore as an Assistant Professor.

  5. Jun 2025

    Defended my PhD in Industrial Engineering at the University of South Florida.

Research

Learning inside optimization algorithms

Hierarchical routing problems require solving many routing subproblems along the way. I train neural networks to predict what those subproblems cost, so that search can skip solving them, and pair learned construction policies with metaheuristics that refine what the policy builds.

Trucks and drones

Mixed fleets in which trucks serve as mobile depots for drones, for humanitarian relief, arc routing and pickup operations. I develop exact methods such as branch-and-price, and hybrid genetic algorithms for instances too large to solve exactly.

Automated warehouses

Deep reinforcement learning for decisions made in real time as orders arrive: how autonomous pickers should route, and when and where mobile robots should recharge.

Fair work in last-mile delivery

How gig-economy delivery platforms can spread work and pay more evenly across their crowdshippers. Giving up 2.5% in cost can improve pay equity by as much as 39%.

Working with me

I welcome invitations for talks and seminars, and opportunities to collaborate. Please get in touch by email.

Students who would like to work with me on mathematical optimization or machine learning should read the notes for prospective students before writing to me.