Active Engineering Computing & AI

Uncovering Hidden Preferences in Route Planning with Data Science

Funder ESRC (part of UKRI)
Recipient organisation University of Liverpool
Funding Not disclosed
Period Mar 2025 — Aug 2028

In plain English

AI plain-English summary

Delivery drivers and logistics planners routinely ignore the most efficient routes suggested by their computers, and this project will use data science to find out why. Current route-planning algorithms optimise for distance, fuel, and deadlines, but they miss the subtle local knowledge that experienced planners bring—knowledge of traffic quirks, driver preferences, or customer-specific delivery quirks. This gap between algorithmic recommendations and human adjustments costs logistics businesses time and money. The project will analyse thousands of manual overrides to algorithm-generated plans, using machine learning to detect patterns in when and why planners deviate. If successful, the research will produce smarter prescriptive tools that blend data-driven efficiency with human expertise. For the transport and logistics sector—which quietly keeps supply chains moving—this could mean fewer missed deliveries, lower fuel costs, and routes that work in practice, not just on paper. The project is applied rather than fundamental science, aimed directly at improving the decision-support systems that fleet operators already use.

View original technical description
Fleet planning in the transport and logistics sector has advanced significantly, improving the efficiency of vehicle operations for deliveries and pickups. Digital technologies have enabled the development of algorithmic tools for route planning that not only minimise distances but also reduce fuel consumption, meet delivery deadlines, and manage vehicle load capacities. However, these algorithms may miss subtle details or "hidden preferences" known to experienced planners, who often adjust algorithmic recommendations based on local nuances and driver inclinations. Understanding why planners deviate from algorithmic routes is crucial for logistics businesses. This project aims to uncover planners' hidden preferences by examining manual adjustments to algorithm-generated plans. By understanding the factors behind these adjustments, the project will enhance existing algorithms with both data-driven insights and human expertise. Using data science, advanced analytics and machine learning, the project will identify patterns, trends, and correlations in datasets, create predictive models that incorporate human insights, and ultimately result in improved prescriptive tools for decision-making in fleet planning.

View the original record at the funder ↗

Researchers

Syed Haider (Student)

Related Research

Grants with similar aims, by meaning.

A scaled and sustainable demand responsive transport service
Development and adoption of an AI solution to improve the utilisation of logistics capacity in transport sector
Artificial Intelligence and Mechanism Design for Optimal Routing of Connected Vehicles
Price Optimisation In Networks for Transport with Signals using Artificial Intelligence (PointsAI)
Optimising Air Transport Route & Demand Planning Using AI Powered Data Analysis

Original classification

Studentship

Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.