RATIONAL - Robust And daTa-drIven Optimization for iNstAnt Logistics

The global aim of RATIONAL is to model and solve realistic transportation problems arising from instant logistics services, focusing on services provided by food delivery companies, which are widely used and constantly expanding. These services face decision-making challenges integral to instant logistics services. Specifically, RATIONAL concentrates on the operational management of on-demand delivery services that utilize crowdsourced delivery. By analyzing actual operational data and considering uncertain factors such as customer demands, delivery times, and operational environment parameters, the project aims to develop data-driven robust optimization methods for crowdsourced capacity, riders’ wages, delivery dispatching, and route optimization at both tactical and operational levels. RATIONAL explores the scientific frontier by combining machine learning with robust optimization methods to provide new decision-making tools and enhance research achievements in related fields with significant scientific implications. RATIONAL will conduct research that straddles the boundary between basic and applied research and aims to achieve the following primary objectives and outcomes: 1) The project seeks to extend the current instant delivery scenario from static and deterministic to dynamic and multi-stage decision-making under uncertainty. The project proposes novel research frameworks for tactical and operational problems combining data-driven thinking, machine learning, and robust optimization methods. It aims to improve the theoretical operations management of real-time dispatching on data-driven food delivery platforms. By utilizing historical data and side information (or feature information), the project will employ machine learning techniques to describe the ambiguous distribution of environmental parameters and establish related robust optimization models. Moreover, the project will employ cutting-edge techniques and conduct fundamental research to design and develop solution algorithms. 2) The project will conduct applied research based on real data and apply theoretical research results to actual operation and management activities. It will address the challenges and problems in real-time dispatching management and aim to improve management performance. The project will provide data-driven, more accurate, and robust decision-making support for on-demand delivery service operations management. It will help relevant enterprises or platforms reduce costs, increase efficiency, improve customer satisfaction, enhance corporate competitiveness, and have significant social and economic implications to Qatar and, more broadly, to the world. 3) The project will lay the foundation for local research groups to explore recent and innovative topics like robust and data-driven optimization, which find applications in manufacturing, sustainability, energy systems, agriculture, etc. Any academic institutions in Qatar do not currently cover these topics.

Hamad Bin Khalifa University (HBKU)