Living Architecture Lab, UCL
Type
Teaching /
Professional
Description
RC3 (The Living Architecture Lab) interrogates the notion of living architecture as a coupling of living systems with the assembly and formation of architecture. The studio holistically reappraises the linear building life cycle, proposing instead adaptive construction with simple flexible parts. The Lab's research focuses on developing autonomously reconfigurable buildings with situated and embodied agency, facilitated variation, and artificial intelligence.
Role
As Researcher and Technical Tutor, Sergio teaches the robotics module and provides weekly desk critiques for master’s graduation projects. He supports students in developing adaptive robotic and computational systems, while developing research on multi-robot design and multi-agent coordination.
Tutors
Tyson Hosmer
Octavian Gh.
Philipp Siedler
Baris Erdincer
Ziming He
Sergio Mutis
Institution
UCL, University College London
Years
2022-
present
Research Themes
Collective Robotic Construction
Multi-Agent Reinforcement Learning
Autonomous Assembly
Adaptive Architecture
Digital Fabrication
Computational Geometry
24-25 Project Highlight
AUTOPOESIS
by Grahesh Bhandari, Elif Solcun, Shengyuan Yu, Jie Qiu, and Emre Tas
A robotic construction system that reuses recycled stone aggregates, dust, and plastics as interlocking modular components, fabricated through voxel-based geometries using robotic carving and 3D printing. Simulation and machine learning evaluate structural and spatial behavior over time, allowing assemblies to adapt and be reconfigured.
Awards:
Bartlett, AD Bronze Medal. 2025. UCL. London, UK.
3D Packing Algorithms
A volumetric agent-based workflow generates a catalog of 26 interlocking component geometries and algorithmically packs them in three dimensions. Local geometric rules and spatial constraints ensure structural stability while enabling dense, reconfigurable aggregation without fasteners.
Material Research
The component geometries are digitally fabricated through robotic carving and additive manufacturing using recycled stone, dust, and plastic composites, and assembled into experimental prototypes up to 1.5 m high. Physical testing evaluates structural behavior, tolerance, and reusability across multiple assembly cycles.
22-23 Project Highlight
RE:SPACE
by Yuechuan Jin, Yongye Xie, Yaosheng Tang, and Weiheng Zhao
An intelligent robotic system that reactivates underused urban buildings by combining automated construction robots with space-generating algorithms. Prefabricated modules and computational planners enable flexible reconfiguration of work and living spaces, supporting sustainable reuse.
Awards:
Bartlett, AD Computation Medal. 2025. UCL. London.
3D Multi-agent Pathfinding
A pathfinding system that coordinates multiple robotic agents, including a wheeled track robot and two-ended crawler, to navigate in 3D space, enabling autonomous navigation and task sequencing for adaptable construction.
Multi-agent Space Generation
The project features a agent-based space generation system where spatial units act as individual agents with learned behaviors, negotiating occupation and growth based on structural constraints and spatial interest.