Explainable AI for Passenger Flow and Commercial Decision-Making in Complex Airport Systems The University of Manchester in United Kingdom

This PhD project at The University of Manchester focuses on advancing explainable artificial intelligence (AI) for passenger flow and commercial decision-making in complex airport systems. Airports represent highly dynamic environments with millions of passengers, hundreds of interacting systems, and real-time operational and commercial decisions made under uncertainty. The project is conducted in close collaboration with Manchester Airports Group (MAG), the UK’s largest airport operator, providing direct access to large-scale, real-world industry datasets.

As a student, you will develop and test innovative methods in probabilistic deep learning, representation learning, graph neural networks, and uncertainty-aware models. These approaches will be applied to MAG’s rich, multi-modal spatio-temporal datasets, including anonymised passenger movement, retail transactions, and operational activity. The research aims to reconstruct and represent noisy, incomplete passenger trajectories, uncover hidden behavioural patterns, link spatial connectivity and flow dynamics, and build explainable models that support trustworthy decision-making. Additionally, you will create data-driven simulation tools to explore how operational changes impact flow, congestion, and commercial performance.

The project offers regular interaction with MAG’s Data Analytics team, including online and onsite meetings, industry supervision, secure data access, and a dedicated MAG laptop. The methodological focus is cutting-edge, with applications extending to transport, logistics, urban mobility, and other geospatial domains. Emphasis is placed on transparency, uncertainty, and human-centred AI, supporting impactful outputs for future airport operations, commercial strategy, and customer experience.

Applicants should have a first-class degree or distinction MSc in computer science, data science, statistics, engineering, geospatial science, management science or a related field. Strong skills in Python or R, machine learning (especially probabilistic or unsupervised methods), and an interest in optimisation and simulation are required. Excellent communication skills and enthusiasm for industry engagement are expected. The selection process is free from bias, and the university actively encourages applicants from diverse backgrounds and career paths. Flexible working arrangements and part-time study options are available to support work-life balance.

This is a fully funded UKRI AI CDT 4-year program, with home tuition fees provided and a tax-free stipend at the UKRI rate (£20,780 for 2025/26). The TechExpert pilot offers stipends of £31,000 for students eligible for home fee status. The start date is September 2026. The project is ideal for students motivated by foundational AI challenges, real-world impact, and industry collaboration. For more information or questions, contact Prof. Julia Handl ([email protected]) or Dr. Ali Hassanzadeh ([email protected]).

To apply, use the University of Manchester application portal, select 'PhD in Artificial Intelligence', and specify the project title and supervisor names. Upload all required supporting documents, including transcripts, CV, supporting statement, and English language certificate if applicable. Ensure referee contact emails are official university/work addresses. Incomplete applications will not be considered.