Optimizing Logistics under Uncertainty

On Thursday 29 January Pankaj Kumar Sinha will hold a trial lecture on a prescribed topic and defend his thesis for the PhD degree at NHH.

Logistics systems quietly shape everyday life by moving goods and people reliably through complex networks. Behind this reliability lies a central challenge: key decisions must be made under uncertainty. Containers arrive unpredictably, demand changes over time, and once a decision is taken—such as accepting a booking or dispatching resources—it cannot be undone.

This thesis explores how logistics systems can remain efficient and trustworthy when the future is unknown.

The research focuses on rail-based freight and passenger transport, drawing on real-world settings in Indian Railways and public transport in Norway. Across three related papers, the thesis studies sequential decision problems where resources are constrained, operational rules are strict, and decisions must be made in real time.

The first paper examines container terminals handling double-stack freight trains in India. It introduces a simple, fast procedure for storing and loading containers that guarantees feasibility without relying on computationally heavy optimization models.

This ensures that once a container is accepted, it can always be loaded as promised.

The second paper looks at how container bookings should be assigned to fixed-schedule trains as they arrive one by one. It combines advanced optimization models with interpretable decision trees, showing how insights from complex models can be translated into clear, practical rules that operators can apply in real time.

The third paper applies the same ideas to passenger transport, developing strategies for deploying ticket inspectors under uncertain passenger flows. The results balance effectiveness with fairness and public trust.

Together, the thesis advances a simple idea: uncertainty is not a flaw to eliminate, but a condition to design for. By combining optimization, interpretability, and operational realism, the research shows how logistics systems can absorb uncertainty while remaining reliable and credible.