MIT researchers have harnessed AI and robotics to upcycle forks in tree trunks for use as sustainable connecting structures in buildings
Tree trunks are the go-to resource for timber in buildings, but researchers in the US have demonstrated how Y-shaped tree sections can replace critical connections in structures, avoiding waste and reducing carbon emissions in the process.
The Digital Structures research group within MIT’s Building Technology Program developed a ‘design-to-fabrication’ workflow using computational tools to match different shaped wooden forks (where a trunk or branch divides in two) with equivalent nodes in an architectural model.
Custom software automates the fork selection and coordinates manufacture by robots, while the final structure is assembled manually. A prototype installed in the MIT lab includes 12 nodes, and plans are afoot to install a larger version with around 40 nodes as an outdoor pavilion in Somerville, Massachusetts.
According to the researchers, tree forks are an ideal replacement for conventional structural connections because the complex internal fibre structure and geometry transfer forces very efficiently.
Caitlin Mueller, associate professor of architecture and of civil and environmental engineering in the Building Technology Program said: 'Tree forks are naturally engineered structural connections that work as cantilevers in trees, which means that they have the potential to transfer force very efficiently thanks to their internal fibre structure … If you take a tree fork and slice it down the middle, you see an unbelievable network of fibres that are intertwining to create these often three-dimensional load transfer points in a tree.'
The approach could support lower-carbon buildings as well as upcycling wood that would normally either be turned into pellets and burned, or ground up to make garden mulch, releasing sequestered carbon into the atmosphere.
The MIT project is part of a four-year research programme into upcycling waste materials for construction and was supported by funding from the School of Architecture and Planning.
The Digital Structures team developed a five-step design-to-fabrication workflow using digital and computational tools to optimise fork structures and minimise cutting to reduce costs. According to Mueller, the approach is potentially scalable for use in industrialised processing systems.
The wood used to build a prototype structure at the MIT lab was sourced from the Urban Forestry Division of the City of Somerville, Massachusetts. An inventory of tree forks was laser scanned in high-resolution, then stored as 3D models in a digital library.
A special metric was used to match Y-shaped nodes in a sample architectural design with tree forks in the library, quantifying how well the geometries aligned to achieve optimal load transfer on to vertical elements.
The overarching aim was to achieve the best overall distribution of tree forks across nodes in the target design. Different fork-to-node distributions were tested using an algorithm to generate an overall matching score and identify the most structurally efficient use of the tree fork inventory.
Researchers found the accuracy of the match improved as the number of forks in the digital library increased. The best scores had around three times as many forks in the library as nodes in the target design.
An element of design flexibility in the software enables designers to alter the overall shape or geometry of the design, while the algorithm automatically recalculates optimal fork-to-node matching. Furthermore, structural analysis checks deflections, strain energy, and other performance measures to ensure integrity.
Moving into fabrication, another custom algorithm computes cuts needed to make a given tree fork fit into its assigned node, using a robotic arm and band saw. The structure is then assembled by hand, matching labels on joints to the correct positions.
Future research plans include the use of larger material libraries, some with multi-branch forks, and the use of computerised tomography scanning technologies that automatically generate detailed 3D models of tree forks, including precise fibre orientation and density.