Skip to content
Skip to navigation menu

Dr Julie Vile

Overview

Miss Julie Williams Position: Research Associate Email: VileJL@cardiff.ac.uk
Telephone: +44(0)29 208 70986
Fax: +44(0)29 208 74199
Extension: 70986
Location: M/1.11

Teaching

Autumn Semester
MAT002 - Statistical Methods

External Courses

An Introductory Course in data management

SAS data step, macro programming and SQL

Research Interests

Operational Research Methods Applied to Healthcare

Queueing Theory

Statistical Methods and Applications

Mathematical Modelling (Specificially for Aneurin Bevan Health Board)

Research Group

Operational Research

Other Duties

Mathematical modeller at Aneurin Bevan Health Board

Publications

 

Research

My research associate post is jointly funded by Cardiff School of Mathematics and Aneurin Bevan Health Board, so my work commitments are accordingly shared between the two organisations. Within the Health Board, I am actively developing a new Mathematical Modelling Unit to explore the potential system impacts of different interventions to ensure optimum benefits realisation. At Cardiff School of Mathematics, my research interests can be divided into three broad streams:

  • Novel statistical and time-series methods that can be applied for interdisciplinary research
  • Stochastic modelling for planning and management of healthcare resources
  • Crew rostering and shift scheduling algorithms to promote efficient resource allocation

Whilst my is research is primarily focussed on healthcare applications, and I have combined the above three techniques to design a workforce capacity planning tool for the Welsh Ambulance Service Trust (WAST), the research methodologies could equally be applied to other service organisations.

Further details of my research contributions are given below:
Novel statistical and time-series methods that can be applied for interdisciplinary research
Projects I have worked on include forecasting future demand to be exerted on the Welsh Ambulance Service using Singular Spectrum Analysis (SSA), and developing minimal detectable change of common tests used by physiotherapists, with particular application to Huntington’s disease.


Stochastic modelling for planning and management of healthcare resources
Particular areas of impact in this area include the development of approximate and numerical time-dependent queueing theory techniques to enable their application within priority service systems. Recognising that healthcare organisations are often closely scrutinized for the quality of service they provide, and expected to achieve strict targets that ensure given proportions of customers are served within acceptable time frames, I have further presented algorithms that compute the time-dependent waiting time distribution in priority systems,  together with the methodology required to provide  minimum staffing requirements.


Crew rostering and shift scheduling algorithms to promote efficient resource allocation
Common IP solvers are not always capable of providing optimal solutions to complex scheduling and rostering problems. I have contributed to the research in this area by formulating the problems as integer linear programs that may be solved using novel heuristic search techniques, without requiring specialist software. The heuristic algorithms can be embedded within a spreadsheet tool, and used to provide feasible and desirable, though not necessarily optimal, shift schedules and rosters.

Biography

Qualifications

PhD – “Time-Dependent Stochastic Modelling for Predicting Demand and Scheduling of Emergency Medical Services” (Funded by LANCS, Cardiff University, 2012)
Bsc – Mathematics, Operational Research and Statistics (First Class Honours, Cardiff University, 2009)

Positions Held

2013-Present: Post-doctoral research associate in Operational Research, Cardiff University/Aneurin Bevan Health Board
2009-2012: Undergraduate tutorial assistant, Cardiff University
2007-2008: Statistical analyst, Welsh Government (Undergraduate Industrial Placement)