ICSS DOCTORAL TRAINING CENTRE LOGO

The Future of Complex Systems Simulation Science

The impact of the DTC has been both broad and fundamental. Our graduating students have moved on to a range of disparate careers, including academic, industrial, consultancy, financial and policy positions. Our graduates now span the globe working on an extensive array of projects from regulating banking systems to advising on oceanographic sustainability, from transport planning to swarm intelligence. Notable destinations (to date) include:

  • Six students staying on at the University of Southampton as part of the EPSRC Doctoral Prize scheme providing a full years research funding to develop as independent scientists.
  • Out of a twenty-four students moving on to postdoctoral research, eight are in overseas institutions, including University of Tokyo, MIT, and Nanyan Technology University.
  • Individual non-academic destinations include TSL Technology Ltd, Quadrature Capital, Ericsson, Decisions Lab UK, Alpha Algorithmics, the Welsh Assembly, and NewForesight.
Background Total
Physics 23 Engineering 20 Computing 20 Biology 10
Mathematics 12 Earth Sciences 12 Politics 3 Chemistry 2
Social Science 2 Law 1 Archaeology 1 Enrolled 106
The disciplinary background experience of the 106 students recruited by the DTC. To date the DTC has a completion rate of 75.5%.
Some DTC students

Systemic Banking Risks - Bob De Caux

Bob De Caux

The DTC gave me the technical experience to pursue an AI career. Previously my focus was in investment banking. This mix of commercial and technical skills is driving my research experience and my subsequent career. A highlight was my involvement with CRISIS, a research project at the University of Oxford under Professor Doyne Farmer. Through CRISIS I showcased my ideas regarding how to model systemic risk in a banking system within the context of an EU project. The CRISIS framework went on to be used by the Bank of England for simulating policy changes, and I was proud to have contributed code and algorithms to it.

During my time at the DTC, I interned at Logical Glue, a machine learning start-up. I continued working for them part-time as a data scientist before becoming full-time Chief Product Officer, in charge of developing the machine learning cloud-based platform. From there I became Chief Executive Officer in 2018 with a now established platform for “explainable AI” in both finance and insurance, helping solve a lot of the regulatory concerns that come from the standard “black box” machine learning approaches.

I am with the software company IFS as their VP of AI and Robotic Process Automation, leading their journey to incorporate AI and machine learning into their software products around the world to make them more effective.

The Dynamics of a Bacterial Plant Pathogen - Melissa Tomkins

Melissa Tomkins On a TrainI use complex modelling methods in order to understand the dynamics of a vector-spread bacterial plant pathogen: phytoplasma. In particular, to investigate how the presence or absence of particular virulence proteins, whose impact on the plant and insect vector behaviour has been well-studied, influences the dynamics of the infection. This is to be achieved through the implementation of a multi-scale, mechanistic model, which includes explicit representation of both the insect vectors, and the plant hosts. My role is to design and code the model, and integrate it with GIS data, in order to investigate the infection dynamics in a real-world scenario.
John Innes Centre, Norwich.

Real World Simulation - Ben Schumann

Ben SchumannSince leaving the DTC, I joined a small simulation consulting firm in London for three years, building simulation models for aerospace and defence clients. Two years ago, I joined McKinsey to push simulation as the perfect consulting tool, helping mining and supply chain clients and creating an internal group of simulation enthusiasts. Recently, I started my own simulation consultancy, building largely process models to support decision making. As you can see, I fell in love with simulations during my PhD and turned it into a career that I enjoy every day.

Transport Simulation - Chris Osowski

Chris Osowski

Since leaving the DTC I have continued to develop my virtual reality and simulator work through self-employment and winning funding to work on a series of projects in partnership with the University. More recently I have returned in-part to a previous employer where now, in addition to putting my transport planning skills to good use (again), I am also helping with the delivery of their new ‘Digital Business Unit’ with my range of skills – built out further through my time in the DTC – and also leading the roll-out of Virtual Reality and related software to wider use in across that company’s full range of engineering disciplines.

Research Software Engineer - Sam Mangham

Sam ManghamI am now employed as a Research Software Engineer at the University of Southampton. My experience with the DTC and my PhD showed me the importance of high-quality, accessible and sustainable research software across a wide range of disciplines. This has inspired me to pursue a career in the new and expanding field of Research Software Engineering, to try and increase the quality of code throughout research.

Sustainable Oceanic Resources - Joe Scutt-Philips

Joe Scutt-Phillips

Having completed a number of postdoctoral positions in Australia and the Netherlands on simulating the movement and distribution of tropical tuna species, I now work for The Pacific Community NGO in Nouméa, New Caledonia, where I provide advice to Pacific Island countries and territories on the sustainable management of their oceanic resources. This work principally centres on analysis of tagging data from tropical tunas, the use of controversial fish aggregating devices by fishers and how it interacts with these fish species, and developing simulation tools for hypothesis testing and experiment design.

Maike Sonnewald - Earth, Atmospheric and Planetary Sciences

Maike Sonnewald is a postdoctoral associate at the Massachusetts Institute of Technology, the University of Texas at Austin and a visiting postdoctoral associate at Harvard University. Dr. Sonnewald earned an M.Sc. in Physical Oceanography (magna cum laude), and a PhD in Complex Systems Simulation and Physical Oceanography from the University of Southampton and the National Oceanography Center. Her research focus is on developing pathways between numerical simulations, data and theory. The intersection between theoretical inference and a fair exploration of numerical tools provide exciting pathways for progress. This is particularly exciting as a means to move oceanographic numerical work into a probabilistic framework, which facilitates direct insight from observational work from a Bayesian statistical framework.

Projects

Maike is involved in a range of modeling studies including the ECCO state estimate, CBIOMES and OCCIPUT from the French MOPGA project. This work focuses on exploring how modeling tools impact our understanding of the underlying dynamics, where Dr Sonnewald has demonstrated the utility of machine learning approaches to understand decadal dynamics. This understanding can be applied to assess future changes under climate change. Her work is targeted at developing understanding of the system's complexity, with the goal of harnessing this insight to understand variability and likelihood of future change.

Lenka Pitonakova

I am currently working on neural-network based self-monitoring algorithms that allow robots to learn about their environment in an unsupervised fashion and to identify new objects and situations. The work is at the University of Bristol and in partnership with Thales, who develop autonomous systems for the transport, space and other industries, allowing us to tackle both academic and applied science questions.

Outreach

The most prominent outreach activity I did was creating a game based on my work in multi-agent systems, The Hive Mind. In the game, player uses a visual interface to program a robot swarm that builds structures on an alien planet. We presented the game at the 9th European Conference on Game-Based Learning (ECGBL 2015) and at the University of Southampton open day in 2015.

  • The Hive Mind game is available for free on http://thehivemind.lenkaspace.net
  • Read more about it and watch videos on http://lenkaspace.net/code/games/theHiveMind

Projects

  • T-B Phase, can be found on http://www.bristol.ac.uk/engineering/research/t-bphase/
  • My PhD and post-doc project, Designing Robot Swarms, can be found on http://designing-robot-swarms.lenkaspace.net/
  • SI have also been developing a python library for plotting, pyCreeper: http://pycreeper.lenkaspace.net/

Computational Models of Human Heart Tissue - Alex Diem

Alex DiemAlex Diem

I am now a Postdoctoral Fellow at the Simula Research Laboratory in Oslo, Norway, working on cardiac perfusion modelling. The DTC helped me to make the choice towards staying within the academic career path and networking connections I was able to make whilst studying with the DTC have helped me find a fantastic postdoc opportunity. Whilst it may seem that my current research topic is very different from my PhD research, the methods and equations I use have largely stayed the same, providing me with the opportunity to demonstrate that my research and methods are applicable to a variety of biomedical problems.

Geological and Planetary Sciences - Stuart Bartlett

Stuart BartlettI am currently a postdoc in the Division of Geological and Planetary Sciences at California Institute of Technology. Broadly speaking, my research focus is the origins of life, and the thermodynamics of complex systems. My PhD and fellowship research focussed on the thermodynamics of convection and reaction-diffusion systems. I still continue to work in those areas, but am also extending previous work to explore whether associative learning can emerge in chemical systems relevant to the origins of life.

Senior Machine Learning Research Scientist -Davide Zilli

Davide Zilli I am currenlty a Senior Machine Learning Research Scientist at Mind Foundry, an Oxford University startup that span out of the Machine Learning Research Group at the Engineering Department. The DTC gave me both the preparation and the academic connections to get a job as a Postdoctoral Research Fellow at Oxford University.

Massimo Stella - Language Modelling

Massimo StellaMassimo StellaAfter submitting my thesis in September 2017, I started a postdoc in Data Science and Health at Fondazione Bruno Kessler, a prominent institution in Italy for artificial intelligence and data science. I worked in the newly created lab of Complex Multi-layer Networks led by Dr. Manlio De Domenico. I learned a great deal about automatic language processing and started gaining experience with machine learning techniques for analysing Big Data related to language and social media, beyond network models.

After one year, I secured a remote consulting job related to language modelling and used that as an opportunity for starting my adventure in data science and scientific consulting. As a complex data scientist and founder of Complex Science Consulting (www.compscicons.net), my research focuses on multi-relational or multiplex network models for detecting patterns in real-world systems such as human cognition and language.

Projects

One of the flagship projects of my new research line includes Complex Forma Mentis (www.complexmentis.com), a data-driven approach for detecting and comparing the students, teachers and researchers' perceptions of STEM subjects as knowledge networks. In 2017, the prototype phase of the project involved 159 students and 59 researchers and the resulting data was presented at the Network Science and Education Symposium 2018 at NetSci18.

Software Engineer at Cyprotex. - James Heppell

James HeppellJames HeppellAfter my PhD, I collaborated with Cranfield University (G. Kirk), Japan International Research Centre for Agricultural Sciences, (M. Wissuwa) and the International Rice Research Institute, (S. Beebout) to address questions related to zinc uptake by rice plants.
I currenlty support the toxicology team at Cyprotex, writing bespoke software to automate and improve their workflow. In addition to maintaining our inhouse software concerning compound registration, assay management and results extraction, I’m leading new improvement initiatives to deal with future workflow demand.

Skyrmion Project - David Cortes

David CortesI'm currently a Research Fellow at the University of Southampton, working as part of a UK based project known as Skyrmion Project. This is a collaborative project of 5 universities in the UK: Cambridge, Durham, Oxford, Warwick and Southampton. The project aims to develop a working magnetic device based on skyrmions and to foster research on magnetic materials capable of hosting skyrmions. More information here: https://www.skyrmions.ac.uk/

Transport Planning - James Snowdon

James SnowdonI've remained working in the same field as my PhD research, but now applying this knowledge and skills professionally.
I spent five years working with SYSTRA, a transport planning consultancy based in Woking, mainly working with Hampshire County Council and Transport for London on transport scheme modelling and appraisal. Aspects included planning a new 'clean air zone' for Southampton, improvements to M3 Junction 9 and smart motorways programme, and development of a new cycling demand forecasting model for London.
In 2018 I moved to Siemens to work jointly with Transport for London and the University of Southampton (including my former DTC supervisor) on the development of a new traffic control system which will be deployed in cities internationally in the coming years.

Senior Research Engineer - Aleksander Dubas

Aleksander DubasAleksander Dubas

Having completed the DTC Complex Systems Simulation programme within the Institute of Complex Systems Simulation. My postdoctoral career began with a Knowledge Transfer Partnership (KTP) between TSL Technology Limited and the University of Southampton. I simulated turbulent flow around and within marine rim-driven thrusters and generators. Unsteady simulations helped to investigate time-variant phenomena such as rotor-stator interaction, while steady-state simulations estimated device performance in a time-efficient manner, allowing iterative optimisation of the design parameters. Embracing the complex nature of the design space allowed the design of a new, world-class, rim driven thruster.
Following the KTP, I continued as a Research Engineer at TSL Technology Limited, developing simulations of downhole equipment involving heat transfer and multi-phase flows.
Now I am a Senior Research Engineer, managing multi-disciplinary engineering projects, a task that the PhD, alongside my wonderful cohort at the ICSS, equipped me for. My research interests remain in the fields of computational fluid dynamics, high-performance computing, turbulence modelling and design optimisation.

EPSRC ICSS