Â鶹´«Ã½ÊÓƵÍøÕ¾

XClose

The Bartlett School of Architecture

Home
Menu

ACADIA Award Win for Bartlett Architectural Design Research

21 November 2019

Bartlett Tutors Tyson Hosmer and Panagiotis Tigas have won the ACADIA Autodesk Emerging Research Award for their paper on autonomous architecture

Speculative autonomous architecture proposal for Mars

Tyson and Panagiotis, both tutors on the school’s Architectural Design MArch, were presented with the award at this year’s for their paper Towards an Autonomous Architecture: Deep Reinforcement Learning for Autonomous Robotic Tensegrity (ART). The paper explores how artificial intelligence can train robotic tensegrity to learn adaptable behaviours related to autonomous mobility, self-structuring, self-balancing, and spatial reconfiguration.Ìý

Tyson and Panagiotis developed a generalisable framework for designing and implementing Autonomous Architecture defined by three key properties: situated and embodied agency, facilitated variation, and intelligence.Ìý

The framework was applied through the development of Autonomous Robotic Tensegrity (ART) using deep reinforcement learning. ART was created together with Bartlett colleagues David Reeves, Octavian Gheorghiu and Ziming He, and Architectural Design MArch students from Research Cluster 3Ìý().

“It is an enormous honour to receive the Autodesk Emerging Research Award at ACADIA 2019, alongside so many inspiring bodies of research from highly respected colleagues and old friends. Our paper establishes and argues for a generalisable framework for autonomy in architecture.

– Tyson Hosmer

°ä´Ç²Ô³Ù°ù¾±²ú³Ü³Ù´Ç°ù²õÌý

Staff
  • Tyson Hosmer, Senior Teaching Fellow, Architectural Design MArch.
  • , Machine Learning Tutor, Architectural Design MArch.
  • David Reeves, Teaching Fellow, Architectural Design MArch.
  • Octavian Gheorghiu, Teaching Fellow,ÌýArchitectural Design MArch.
  • Ziming He, Teaching Assistant, Architectural Design MArch.
Students

Architectural Design MArch students, 2017/18

  • Ziming He
  • Linlin Cao
  • Shi Ren
  • Ning Wang

Architectural Design MArch students, 2018/19

  • Chi Siyu
  • Meng-Yi Lin
  • Jiachen Lei
  • Shiyu Kan

More information

is an international network of digital design researchers and professionals. We facilitate critical investigations into the role of computation in architecture, planning, and building science, encouraging innovation in design creativity, sustainability, and education.

took place at the University of Texas on 24 – 26 October 2019. The theme of the conference was Ubiquity and Autonomy.Ìý


Images

Lead:ÌýSpeculative autonomous architecture proposal for Mars

Other:Ìý
1.2.&3.Ìý2.2m Diameter Tensegrity Robot
4.ÌýReinforcement learning simulation:ÌýAgent Self-Balancing TrainingÌý
5.Ìý2mx1m Tensegrity Assembly Model