At the entrance of the Swiss Tech Cluster known as ‘Zug’ stands ‘Samiaramis,’ a new 22.5-metre superstructure designed by ETH Zurich and Gramazio Kohler Research. The installation uses cross-laminated timber panels with robots to assemble the plates.
ETH Zurich is the studio behind ‘The Vaugirard Social Housing precinct ‘clad in cross-laminated timber, steel and concrete and has designed the mass timber-inspired vertical garden with the help of AI and machine learning.
Construction is increasingly deploying AI and robots to develop 3D and even 4D prefabricated construction materials, with the latest research suggesting that automation could lead to a 50% reduction in material consumption, 63% reduction in emissions and 75% in time-saving.
Using interactive computational design, machine learning and digital fabrication, the architects created a towering superstructure consisting of eight wooden pods suspended by eight steel pillars.
But it is its use of machine learning, thanks to a collaboration with the Swiss Data Science Centre, which is attracting attention.
According to the US-based Construction Specifier, the architecture and design community is increasingly using machine learning to create the buildings of the future.
The project used machine learning to determine the most efficient design variations considering factors like sunshades, rain protection and even plantable surfaces.
The collaboration with the Computational Robotics Lab led to the creation of a custom tool which enables the optimisation of the complex geometry of individual pods while accounting for material properties and fabrication parameters.
Doing this ensured that each panel maintained its flatness and adhered to specified size constraints while enhancing structural integrity.
The cross-laminated timber plates were assembled using a robotic procedure developed by ETH Zurich in collaboration with Timbatec Timber Construction Engineers in Switzerland.
This process enables the butt joint bonding of wood, producing large areas of intricately folded wood structures.
Lightweight timber structures are suitable for automated on-site assembly due to the high precision in prefabrication and the low component weight.
According to research from Goldman Sachs, architecture and engineering are among the top 3 industries impacted by the push to AI and automation, with 37% of roles disrupted by new technology.
In July, Wood Central reported that leading architectural practices use “Midjourney” software to expedite the conceptual design phase.
The software presents project-specific design precedents that align well with client requirements, visions, and aesthetic preferences.
The software is already used for various projects, including mixed-use mass timber buildings.