Paul Shepherd

Paul Shepherd is Lecturer in Digital Architectonics. His research involves the development of new computer-based methods and tools for both architects and engineers to help with the creation, optimisation and realisation of complex geometry buildings.

Paul studied Mathematics at Cambridge before moving to Sheffield to complete a PhD in the Department of Civil & Structural Engineering. His work in the Fire Research Group involved developing highly non-linear finite element modelling techniques. Paul also collaborated with the University of Ulster on a parallel project involving a series of physical tests on restrained columns in fire and attended the Cardington tests, where a full-scale 8-storey building was repeatedly set on fire.

On completion of his PhD, Paul joined the Special Structures group of Buro Happold. He began by re-writing their in-house fabric structure analysis software and then using his program on structural engineering projects, particularly those involving fabric, or with complex geometry. He developed a special interest in the use of Parametric Modelling in the design process and went on to set up a specialist software and analysis group within the company, which used programming and mathematical modelling techniques to help develop solutions to complex engineering design problems. Some of the landmark projects Paul has worked on include the Stuttgart21 Rail Station, the Japan Pavilion for the Hanover2000 Expo and more recently the Emirates Stadium at Arsenal and the new Aviva Stadium at Lansdowne Road in Dublin.

Paul joined the University of Bath as a Research Fellow in 2007 and is now Lecturer in Digital Architectonics and the Director of Studies for the associated MPhil. He is also a passionate science communicator and has a long track record in outreach work promoting Maths and Engineering to the public, and especially to high-school students.

Paul’s research involves the development of new computer-based tools for both architects and engineers to help with the creation, optimisation and realisation of complex geometry buildings. Until very recently, computers had been used almost exclusively by the building industry as a form of electronic paper, helping designers to communicate their ideas, but not actually contributing to the design process itself. But with the recent development in computational speed, graphics and networking, this means that the full potential of the computer is being wasted.

By combining mathematical techniques such as Dynamic Relaxation, Multi-Objective Optimisation, Parametric Modelling and Surface Subdivision, Paul aims to facilitate the design of complex geometry forms, using computational techniques to improve the potential structural and environmental performance and constructability of the resulting shapes. He hopes that computers will become active participants in the design process, as well as providing a common platform on which all members of the design team can interactively share ideas and truly collaborate to create innovative and sustainable buildings.