Autonomous Robotic Construction and Assembly

Lead Research Organisation: University College London
Department Name: Computer Science


Numerous manufacturing sectors are adopting advanced forms of computation and robotics to develop autonomous production logistics frequently referred to as 'Industry 4.0'. Construction offers significant challenges compared to other manufacturing sectors due to the fluctuating environmental and circumstantial conditions on each site that require more robust and adaptive means of autonomy. UCL-CS's Autonomous Manufacturing Lab (AML) develops autonomous manufacturing software platforms that engage in robotic production logistics and the integration of both explicit and generative design engineering, with a specialisation in Construction. The iCASE PhD research will leverage this in development of advanced forms of robotic assembly that will involve the development of novel uses of computer vision and machine learning in order to provide
learning and cognition capabilities within construction robotics. These enhanced capabilities will facilitate novel forms of human-robot collaboration and enable construction to be sensitive to variations in on-site conditions, material dynamics and supply chains logistics.

The doctoral research will aim to develop an adaptive building assembly software that will be demonstrated through a robot platform (Eg. one of: 6 axis industrial robots, scale-able cable robots or multi-axis snake robots) that endows self-awareness and sensor-feedback capabilities in order to develop real-time adaptive design-engineering and production logistics for novel building assemblies. This will involve a number of physical demonstrations that will be programmed as research output milestones throughout the PhD research period. Examples include: large scale self-supporting Catalan vaults, complex and intricate timber or steel space-frames/shells, or prefabricated modules that involve extreme tolerances and dexterous assembly sequences not able to be performed by human workers. The research will involve the integration of a generative design methodology that incorporates structural and environmental concerns (such as the minimisation of material and energy) whilst providing novel design outcomes that are integral to the robotic manufacturing methods explored. The research will address the lack of context and material sensitivity within robotic construction through the development of bespoke means and use of computer vision and machine learning in order to obtain material, semantic, positional and contextual awareness for behaviour-based decision making.


10 25 50

Studentship Projects

Project Reference Relationship Related To Start End Student Name
EP/R512400/1 01/10/2017 31/03/2022
1934295 Studentship EP/R512400/1 26/09/2017 11/04/2022 Julius Sustarevas