UCL Discovery
UCL home » Library Services » Electronic resources » UCL Discovery

Robotic Training for the Integration of Material Performances in Timber Manufacturing

Brugnaro, Giulio; (2020) Robotic Training for the Integration of Material Performances in Timber Manufacturing. Doctoral thesis (Ph.D), UCL (University College of London). Green open access

[thumbnail of 201113_Brugnaro_Robotic Training for the Integration of Material Performances in Timber Manufacturing.pdf]
Preview
Text
201113_Brugnaro_Robotic Training for the Integration of Material Performances in Timber Manufacturing.pdf - Accepted Version

Download (76MB) | Preview

Abstract

The research focuses on testing a series of material-sensitive robotic training methods that flexibly extend the range of subtractive manufacturing processes available to designers based on the integration of manufacturing knowledge at an early design stage. In current design practices, the lack of feedback information between the different steps of linear design workflows forces designers to engage with only a limited range of standard materials and manufacturing techniques, leading to wasteful and inefficient solutions. With a specific focus on timber subtractive manufacturing, the work presented in this thesis addresses the main issue hindering the utilisation of non-standard tools and heterogeneous materials in design processes which is the significant deviation between what is prescribed in the digital design environment and the respective fabrication outcome. To begin, it has been demonstrated the extent to which the heterogeneous properties of timber affect the outcome of the robotic carving process beyond the acceptable tolerance thresholds for design purposes. Resting on this premise, the devised strategy to address such a material variance involved capturing, transferring, augmenting and integrating manufacturing knowledge through the collection of real- world fabrication data, both by human experts and robotic sessions, and training of machine learning models (i.e. Artificial Neural Networks) to achieve an accurate simulation of the robotic manufacturing task informed by specific sets of tools affordances and material behaviours. The results of the training process have demonstrated that it is possible to accurately simulate the carving process to a degree sufficient for design applications, anticipating the influence of material and tool properties on the carved geometry. The collaborations with the industry partners of the project, ROK Architects (Zürich) and BIG (Copenhagen), provided the opportunity to assess the different practical uses and related implications of the tools in a real-world scenario following an open-ended and explorative approach based on several iterations of the full design-to-production cycle. The findings have shown that the devised strategy supports decision-making procedures at an early stage of the design process and enables the exploration of novel, previously unavailable, solutions informed by material and tool affordances.

Type: Thesis (Doctoral)
Qualification: Ph.D
Title: Robotic Training for the Integration of Material Performances in Timber Manufacturing
Event: UCL (University College of London)
Open access status: An open access version is available from UCL Discovery
Language: English
Additional information: Copyright © The Author 2020. Original content in this thesis is licensed under the terms of the Creative Commons Attribution 4.0 International (CC BY 4.0) Licence (https://creativecommons.org/licenses/by/4.0/). Any third-party copyright material present remains the property of its respective owner(s) and is licensed under its existing terms. Access may initially be restricted at the author’s request.
UCL classification: UCL
UCL > Provost and Vice Provost Offices
UCL > Provost and Vice Provost Offices > UCL BEAMS
UCL > Provost and Vice Provost Offices > UCL BEAMS > Faculty of the Built Environment
UCL > Provost and Vice Provost Offices > UCL BEAMS > Faculty of the Built Environment > The Bartlett School of Architecture
URI: https://discovery.ucl.ac.uk/id/eprint/10114936
Downloads since deposit
100Downloads
Download activity - last month
Download activity - last 12 months
Downloads by country - last 12 months

Archive Staff Only

View Item View Item