SUPERSPACE is Woods Bagot’s design research group pioneering data driven spatial systems for planning and design. Our team of multi-disciplinary design researchers integrates big data, spatial cognition, behavioural analysis and machine learning to design evidence-based user-centered environments. Digital spatial systems are designed and built in-house through professional software development.

SUPERSPACE’s unique and rigorous approach is founded on an Open Framework for Spatial Simulation developed over more than a decade and recognised with a Royal Institute of British Architects (RIBA) President’s Award for Research in Practice.

We have a long track record of teaching, conducting workshops and producing fundamental R&D with or for partners such as universities, government and private clients. Past partners include: European Union, Fraunhofer Institute Germany, Canadian Social Sciences and Humanities Research Council, UK Higher Education Investment Fund, University College London, ETH Zurich, Science Po Paris, University of Manchester, Columbia University New York, Technical University Munich, UdK/ TU Berlin, Sheffield University, Politecnico di Milano, University of Sydney.

Simulation of Spatial Systems

Designers, artists, societies produce environments by intuition, learned procedures and tradition. We analyse the dynamics and components, specify their relations and encode models that simulate systems that make up such an environments. The digital simulations enable our partners and clients to document their systems, improve their designs or spaces and predict effects of future applications of their designs.

User-Centric Design Systems

Most ‘good’ design results from a clear process based on conceptual objectives. We help students to explore concepts based on people – their perceptions of place, qualities of space, behaviours and relations to others – by introducing them to the quantification techniques that lead to social and spatial data analysis. Design principles and rules are specified from data visualization and pattern mapping and translated into algorithms for a hybrid (digital-physical) design process. Students develop a design systems that produce design instances of the system as an inside-out spatial generator.

Urban Data and Computational Planning

For many cities around the world and at lower scales of neighborhoods there is no or very little data. We instruct partners or collect qualitative data on sites through field studies and analyze it to produce insights for planning. We are particularly interested in social and activation data to produce ethnographic mappings from which planning guidelines about liveability can be generated. Using those guidelines we create bespoke planning simulations to generate novel and citizen centric spatial frameworks.

Data Analysis and Visualization

Analyzing and visualizing big data sets is not trivial. Many partners and clients have lots of data but are inexperienced with analysis or not clear what to look for. We explore large data sets and organize them into clear relationships and hierarchies before designing bespoke real-time and interactive visualizations according to objectives that we workshop with our partners.

Machine Learning for Space Planning

Intuition and perceptions are learned by people. Alternative ways of understanding spatial environments are possible through machine learning and open up surprising insight. We have been pioneering models of machine learning for nearly 20 years in academia and practice in order to provide the field of space planning with new descriptions of spatial perception, building organization and urban conditions.