Please upgrade your browser

For the best experience, you should upgrade your browser. Visit our accessibility page to view a list of supported browsers along with links to download the latest version.

Funded Collaborative Doctorate: Defining Architectural Typologies Through Structural Topologies and Machine Learning

The London Arts and Humanities Partnership (LAHP) are pleased to announce an AHRC-funded Collaborative Doctorate between the Royal College of Art and AKT II (Consulting Structural and Civil Engineers). This studentship will commence in October 2019 and is fully funded for three and a half years, covering tuition fees and stipend. The successful candidate will have full access to the LAHP Doctoral Training Partnership development activities and networking opportunities.

Applications close: 31 May 2019. Apply here.

Project Overview

Many buildings are in functional terms not defined by formal, but organisational and structural typologies. Especially structural design significantly determines both the external form and internal layout of a building. The potential of topological structural models to define more holistic architectural typologies and their implications for new construction and design processes that work across the scales of the building to that of the room or building element and component has been little researched. In order to undertake this research, however, a continuing divide between architects and structural engineers has to be overcome. Thus, the research asks: Can integrating a structural and architectural design process delineate new architectural building typologies that are defined by structural topologies, and do these provide useful design responses to social and economic needs in the building industry? Are these new architectural/structural models more effective when designing adaptable buildings required by increasingly changing building lifecycles and user demands? And, what are the performance parameters that define these typologies and topologies?

Working closely with the Laboratory for Design and Machine Learning at the School of Architecture, RCA, and, a computational research team at AKT II, this practice-led PhD aims to develop new design models and processes with practical applications for buildings that dynamically change their behaviour through mechanical or formal adaptation. It will develop, adapt, and refine algorithms for use in the analysis and design of the built environment through machine learning, with the aim to improve how buildings can adapt to environmental, social, and economic changes, use less energy, and provide better environmental comfort and space efficiency than static buildings. This causality between environmental impacts and built environments is of great societal, scientific and economic interest.

This PhD by Project will use an interdisciplinary methodology that brings together existing, statistical and computational methods of analysis from engineering and the sciences and typological design and research methods from an architectural and arts context. The mixed research methods include a qualitative longitudinal study of architectural typologies and structural topologies, and a quantitative analysis of performance-based design models and processes. Especially the use of rigorous statistical methods through machine learning algorithms, provides novel ways to analyse and inform emerging spatial and performance-based design processes. The proposed methods will bring quantifiable, scientific evidence to architectural design, which is still largely justified as a qualitative problem. One aim of this PhD is therefore to explore how empirical design processes challenge traditional assumptions of efficiency and spatial performance over time.

Research will include a pilot study to test parameters of typological/topological machine learning models, before determining a research focus on one particular architectural/structural typology and build up a statistically significant dataset to train a machine learning algorithm capable of deriving and evaluating an adaptive design model and process. The work will consequently engage critically with practice-led problems, but also question how practice is conceptualised, hereby particularly examining the limitations or possibilities that machine learning brings to spatial design processes, whose application and study is still largely in its infancy in this field.

Websites: and 


Dr Sam Jacoby (School of Architecture Research Leader, RCA)
Alessandro Margnelli (Technical Director, AKT II)
Dr Dragos Naicu (Computational Design Engineer, AKT II | 


Competitive candidates will have advanced computational skills (with a computer science background or equivalent experience highly desirable) and a degree in architecture or structural engineering. 

Studentship details

The AHRC-funded London Arts and Humanities Partnership (LAHP) brings together eight leading British research universities: King’s College London, London School of Economics and Political Science, Queen Mary University London, Royal Central School of Speech and Drama, Royal College of Art, Royal College of Music, School of Advanced Study and University College London. 

The studentship includes a stipend at Research Council UK 2019/20 rates of £17,009 per annum (plus fees at home/EU rates) for three and half years. The awarded candidate will also be entitled to a £550 per annum stipend top-up. As a LAHP student, the successful candidate will have full access to the LAHP Doctoral Training Partnership development activities and networking opportunities, joining a cohort of about 90 students per year. Studentships can be either full or part-time. 

Applicants should have a good undergraduate degree in a relevant discipline, and a Master's-level qualification or equivalent which meets AHRC requirements for research training. Applicants with relevant work/professional experience who are considering doing a PhD are also encouraged to apply. 

Closing Date: 31 May 2019

For more information and to apply for the studentship, please see the LAHP website or email [email protected]