In both nature and society, distribution networks are fundamental, facilitating the exchange of materials, energy and information. As systems evolve, these networks become complex leading to fragile systems at significant risk of failure. Rail networks are a classic example of this, where timetable pressures amplify the effects of mechanical failure, spikes in demand and adverse weather.
The issue of maintaining the integrity of a network is central to all large-scale computer networks operated by the likes of Google, Microsoft, and Alibaba. In tandem with computer scientists, these organisations have created machine learning methods that make these networks resilient to both disturbance and decay. This project will research how these embedded decision making methods can be used to make rail networks resilient.
The project will be a collaboration between the School of Computing and Communication and the Lancaster Environment Centre at Lancaster University, and Digital Rail. It aims to produce a new approach to predict and minimise complex failures within distribution networks (road, rail, air) more generally. As a result, we envisage being able to provide a research environment that is both intellectually challenging and of significant practical relevance, with opportunity to develop meaningful links within the rail sector.
This PhD studentship covers full UK/EU fees and a stipend of £14,777 p.a., in addition to a training grant of £2500 p.a.
Our ideal candidate would be able to demonstrate they can bring energy and enthusiasm to the project, along with a foundation in the kind of technical and numerical skills that such a project necessarily requires. They will have a minimum of 2:1 in their first degree in a STEM related subject and be able to articulate clearly why research is the appropriate next step in their career.
All enquiries to Dr Andrew Jarvis (firstname.lastname@example.org).