A study by MIT researchers suggests that using robotically assembled building blocks could offer a more sustainable and efficient method for constructing large structures compared to some current techniques. The team evaluated the feasibility of building with “voxels,” modular 3D units that form complex, strong structures. They developed three new voxel designs, a robotic assembler, and an interface for creating voxel layouts and directing the robots.<\/p>\n
The researchers found that this voxel-based robotic assembly method could cut embodied carbon \u2014 the total carbon emissions from building materials \u2014 by up to 82% compared to methods like 3D concrete printing, precast modular concrete, and steel framing. The system is also competitive in cost and construction time, though the material choice affects the carbon footprint and cost of the voxels. Further research is needed to assess scalability, durability, fire resistance, and other factors before widespread use.<\/p>\n
Miana Smith, MIT graduate student and lead author, expressed excitement about using robotic assembly of discrete lattices to enhance digital fabrication in construction. Smith collaborated with Paul Richard, a graduate student from \u00c9cole Polytechnique F\u00e9d\u00e9rale de Lausanne, Alfonso Parra Rubio from MIT, and senior author Neil Gershenfeld, MIT professor and director of the Center for Bits and Atoms (CBA). Their findings were published in Automation in Construction.<\/p>\n
Over recent years, CBA researchers have been working on voxels, which are lattice-structured blocks designed for high strength and stiffness, similar to airplane wings and wind turbine blades. “We are applying aerospace principles to buildings,” said Gershenfeld, referencing past work with NASA, Airbus, and Boeing. The researchers assessed the mechanical and environmental performance of existing voxel designs before creating three new geometries for easier robotic assembly.<\/p>\n
The new voxel designs, based on a high-strength octet lattice, allow for mechanical self-alignment into rigid structures. “The interlocking nature of these voxels provides strong mechanical properties without many connectors, speeding up construction,” Smith noted. They devised a robotic system using inchworm-like robots, called MILAbots, which use grippers to place and interlock voxel blocks efficiently.<\/p>\n
The team analyzed the embodied carbon of their designs using plastic, plywood, and steel materials and compared the costs and speed of their robotic system with traditional methods. They found that while plastic voxels performed poorly environmentally, steel and wood offered significant benefits. Steel voxels produced 36% and 52% of the embodied carbon required for 3D concrete printing and precast concrete, respectively, while plywood required only 17% and 24%.<\/p>\n
Projected on-site assembly times for steel and wood voxels averaged 99 hours, compared to 155 hours for existing methods. Although a single MILAbot is slower, a team of 20 working together matches or surpasses current automation methods at lower cost. “This method is incremental, allowing for easy expansion or modification,” Gershenfeld explained.<\/p>\n
The team also created an interface for designing voxel structures and instructing the MILAbots. Looking ahead, they plan to test the system on a larger scale in Bhutan, using a “super fab lab” to replicate robots for a sustainable city project. Future research will focus on stability under lateral loads, improving design tools, enhancing MILAbots, and evaluating voxels with integrated features.<\/p>\n
“Our work supports the practicality of distributed robot assembly for digital fabrication in construction,” Smith stated. The MIT Center for Bits and Atoms Consortia partially funded this research.<\/p>\n