Artificial evolution with miniature robots

​​Evolutionary robotics is a vibrant new sub-discipline of robotics which looks to adapt the processes which shape life on earth to the development of novel engineering systems. This has the potential to hugely enhance robotic sensing, mobility and durability in ways which are economical with energy and material (most animals greatly outperform robots and do so using only a fraction of the energy), not to mention unlocking new and subtle innovations that human design overlooks.  

​ ​Currently evolutionary robotics approaches typically rely on simulation [1], which oversimplifies the mechanics of a problem and makes complex physical phenomena (e.g. fluid flow) out of reach. Where evolutionary approaches include automated manufacture and testing, each design iteration typically requires >6 hours per prototype (1-4 designs per day) and relies on modular assembly approaches which limit the range of designs that can be explored [2]. 

​ ​This project will evolve physical robots 100 times faster than existing approaches, by focussing on miniature systems that can be built and tested rapidly. We will leverage cutting-edge rapid prototyping to produce, test and evaluate hundreds of designs per day with full automation of the entire process, producing vast experimental datasets. The evolutionary experiments will target complex phenomena such as flight, airflow measurement and granular material interaction, for which numerical simulation is intractable. 

​The core hardware and software has already been prototyped and tested at the University of Surrey. The studentship will focus on refining the approach and enhancing the underlying evolutionary algorithm, by adapting the system to solve several different engineering problems with direct application in environmental monitoring, robot mobility and advanced flight. Work will take place in a newly outfitted robotics lab within the School of Mechanical Engineering Sciences, with additional access to wind tunnels, indoor flying spaces and advanced manufacturing to support robot experiments. 

​ ​[1] Cheney, Nick, et al. "Unshackling evolution: evolving soft robots with multiple materials and a powerful generative encoding." ACM SIGEVOlution 7.1 (2014): 11-23. 

​ [2] Hart, Emma, and Léni K. Le Goff. "Artificial evolution of robot bodies and control: on the interaction between evolution, learning and culture." Philosophical Transactions of the Royal Society B 377.1843 (2022): 20210117.​ 

Open to both UK and international candidates.

Up to 30% of our UKRI-funded studentships can be awarded to candidates paying international rate fees. Find out more about eligibility.

You will need to meet the minimum entry requirements for our PhD programme.

​​The project would suit a student from any science/engineering discipline with a strong foundation in programming (Python/MATLAB preferred but not required), and a passion for novel robotics. Beneficial prior experience could include (but is not limited to): mechanical design, electronics, machine learning, rapid prototyping, fluid mechanics/aerodynamics, evolutionary biology. Students should have strong communication skills and a collaborative approach to their work.​