Rian Hughes


A physics PhD student at the University of Oxford, my academic research is on the physics of biological fluids.

I also have an interest in AI, particularly in using human data in VR to train AI agents.

See below for examples of the projects I’ve worked on and if you’d like to find out more feel free to get in touch.


Contact

Connect with me on LinkedIn, check out some of my code on GitHub, or send me an email at rian.hughes@physics.ox.ac.uk


Highlighted Projects


Biography

As an alumnus of the All Ireland Scholarship, I completed my BSc. at National University of Ireland, Maynooth, in Theoretical physics and Mathematics. I then went on to complete my MASt (Math Part III) at the University of Cambridge in Theoretical Physics, with a focus on fundamental particle physics.

I am currently a Marie Skłodowska Curie Fellow reading for a PhD at the Rudolf Peierls Centre for Theoretical Physics at Oxford University. My research interests lie in computational bio-physics, ultracold atom traps, and artificial intelligence.

My PhD research is based on simulating active nematic matter, a class of fluids which inject energy into the system at the particle level; taking inspiration from biological systems, ranging from collective flows in suspensions of micro-tubules to collective chemotaxis.

Outside of my PhD I am an active member of OxAI. One of the projects I’m working on is to build a framework, in which AI agents are trained in social virtual reality environments.
This combines existing platforms such as Neos and Unity’s ML-Agents to leverage human-data in VR, training AI agents to perform a range of tasks. Ultimately, this will also provide insight into human-human interactions.

I thoroughly enjoyed my time at Kraken, a crypto-currency exchange, where I worked as a consultant to develop trading algorithms in a high-security environment.

Education

  • University of Oxford
    PhD Student
    Computational / Theoretical Physics
  • University of Cambridge
    MASt
    Math Part III
  • National University of Ireland, Maynooth
    BSc
    Theoretical Physics and Mathematics

Experience

  • Research Publications.
  • Programming.
    Regular use of C++, Matlab, Mathematica in my PhD.
    Regular use of C# for the OxAI project and Unity.
    Extensive use of Python during my time at Kraken.
  • GitHub

Projects & Publications

AI in Virtual Social Environments

The aim of this project is to build a framework to train AI agents in social virtual reality (VR) environments.
The main motivations are to explore how social interactions with humans can benefit the development of more human-like artificial intelligence.

This idea builds on research in developmental robotics, reinforcement learning, and psychology. However, research in human-AI social interaction has been limited by the use of expensive robots and sensing systems. This project aims to remove these barriers by using new VR technologies.

Programming Skills Involved : C#, API.

Reconfigurable flows and defect landscape of confined active nematics

Active (nematic) fluids can be defined as fluids where the constituent particles actively push on their surroundings, which gives rise to dynamics akin to swarming in bacterial clusters. They have a characteristic length-scale ‘L‘ which determines the scale at which the flows change.
This paper shows that the ratio of L to the confinement size (e.g. channel width) determines what type of fluid flows arise. If the ratio is small (large channel widths) then the flows are turbulent, however as the ratio decreases a range of temporally ordered fluid flow states arise.

Programming Skills Involved : C++, Matlab.

Active Stresses can Govern Collective Chemotaxis

Cells live in a complex environment filled with many cell types, polymers, fluids, etc. It’s important that cells are able to navigate such a chaotic environment, e.g. immune cells in the body need to know where to go to fight off infection.
One mechanism to help with this task is known as ‘chemotaxis’, where cells move towards, or away from, a chemical source. This paper proposes a new physical mechanism for cell clusters to move towards a chemical source.

We model cells as an active (nematic) fluid coupled to a chemical field which regulates the strength of the fluid flows. Active stresses give rise to radial anchoring at the interface, which directs flows outwards from the clusters core. The chemical coupling then provides the necessary symmetry breaking to allow the cluster to move in a given direction.

Programming Skills Involved : C++, Matlab.

Multiple-radiofrequency dressed adiabatic potentials

Remarkably, we are able to not only cool atoms down to near absolute zero, but we can also trap them in mid-air using magnetic fields.

In this paper we develop an analytical approach to sculpt trapping potentials by adding additional electro-magnetic (R.F.) fields. The (non-adiabatic) loss-rate due to the atoms motion is also measured for a double-well potential, both with and without gravity.

Programming Skills Involved : Mathematica

Trading algorithms

Working as a consultant for a crypto-currency exchange, I built trading algorithms to provide liquidity on both liquid and illiquid markets.

This was done using in depth knowledge of Python, APIs, cloud computing and running code in a high security environment.

Programming Skills Involved : Python, API, Cloud Computing, Cryptography .


Blog Posts

Here I discuss a project on training agents to navigate mazes, using Unitys ML-Agents.
Programming Skills Involved : C#.


Contact