13105 - Research Associate in Physics-Informed Machine Learning for Crowd Dynamics

University of Edinburgh; City of Edinburgh; On-site; GBP 41,000 - 49,000; Full time

Job summary

A leading academic institution in Edinburgh is looking for a Postdoctoral Research Associate for an EPSRC-funded project focused on real-time crowd dynamics simulation. The role involves developing a machine learning pipeline and collaborating closely with industry partners and a PhD student. Ideal for those with a strong background in engineering or related fields, this position offers significant opportunities for mentorship and real-world application in public safety and urban planning.

Qualifications

• PhD (or near completion) in a related field.
• Strong expertise in machine learning.
• Solid understanding of mathematical modelling.
• Proficiency in programming languages like Python, Fortran, C++.

Responsibilities

• Contribute to the design and development of a crowd dynamics simulator.
• Develop a physics-informed machine learning pipeline.
• Collaborate with a PhD student and provide mentorship.
• Engage with industry partners for impactful research.
• Document methodology and contribute to publications.

Skills

Machine learning

Scientific computing

Mathematical modelling

Python

Fortran

C++

Analytical skills

Problem-solving

Communication skills

Education

PhD in Engineering, Physics, Applied Mathematics, Computer Science or related field

Job description

Overview

Grade UE07 £41,064 - £48,822 per annum
College of Science & Engineering / School of Engineering / Institute for Multiscale Thermofluids / Machine Learning, Computational Engineering, Crowd Dynamics
Full-time: 35 hours per week
Fixed Term dates: from 1st March 2026, for up to 36 months

The Opportunity

We are looking for a talented, creative, and experienced Postdoctoral Research Associate to join the EPSRC-funded project FLOCKS (Fluid dynamics-Like Open-source Crowd Knowledge-driven Simulator).

Designed in close collaboration with industry leaders, FLOCKS aims to create the world's first real-time, open-source simulator of large, dense crowd dynamics. The simulator will have applications in public safety, urban planning and event management. The research will focus on developing a physics-informed machine learning pipeline to derive governing equations and boundary conditions for macroscopic crowd models from synthetic and real-world data. Close collaboration with a dedicated PhD student, who is developing physics-based models and generating synthetic datasets, will fuel the machine learning framework while also offering a valuable opportunity for mentorship. Thanks to its partnerships with world-leading experts in crowd safety engineering and open-source software development, the project will have a direct impact on real-world applications relating to public safety, urban planning and event management. A final demonstrator will simulate iconic local events (e.g. Hogmanay on Princes Street, an Edinburgh derby football match, or a Murrayfield Stadium concert) using pre-captured datasets to demonstrate the simulator's predictive power and direct relevance to these applications. This is an excellent opportunity for an experienced researcher interested in machine learning, mathematical modelling, and complex systems.

Responsibilities

• Contribute to the design and development of a real-time, open-source crowd dynamics simulator.
• Develop a physics-informed machine learning pipeline to infer governing equations and boundary conditions for macroscopic crowd models from data.
• Collaborate with a PhD student on physics-based models and synthetic data generation, and provide mentorship.
• Engage with industry partners and contribute to impact-focused research outcomes in public safety, urban planning and event management.
• Document methodology, reproduce results, and contribute to open-source software and publications.

Qualifications

• PhD (or near completion) in Engineering, Physics, Applied Mathematics, Computer Science, or a related field.
• Strong expertise in machine learning and scientific computing.
• Solid understanding of the mathematical modelling of physical systems.
• Proficiency in scientific programming (e.g., Python, Fortran, C++).
• Strong analytical, problem-solving, and communication skills.