Computational Design and Co-Robotic Fabrication for Material Reuse in Architecture

As global urbanization accelerates, the construction industry's demand for raw materials continues to rise. According to the United Nations, by 2050, 68% of the global population will reside in cities, creating immense pressure on housing and infrastructure demand. The traditional construction model relies heavily on raw material extraction, exacerbating resource depletion and generating significant construction waste. In recent years, the construction industry has begun exploring circular economy models, seeking to reduce reliance on raw materials, lower carbon emissions, and extend material lifespans through material reuse. Timber, as a primary structural material, is a focal point for material reuse research due to its carbon sequestration capabilities and renewability.

Achieving material reuse in construction presents several challenges. First, heterogeneous reclaimed material inventories complicate design and construction processes. Second, traditional design-to-fabrication workflows cannot effectively address material uncertainties and inventory constraints. Additionally, existing digitized inventory methods often require significant time and cost, making them impractical for large-scale material stocks. Thus, developing an efficient integrated framework to handle reclaimed material uncertainties and achieve efficient material reuse is a pressing issue.

The core innovation of this paper lies in proposing an integrated framework that combines data-driven computational design with adaptive human-robot collaborative fabrication. • Data-driven computational design: Conducts design exploration by considering material inventory lengths and geometries. • Adaptive human-robot collaborative fabrication: Handles material uncertainties through perception-driven material selection and adaptive assembly. • Integrated design-to-fabrication workflow: Achieves seamless integration from design to fabrication, reducing rework and translation errors typical in traditional workflows.

• �� Data-driven computational design: Implemented in Rhinoceros 3D and Grasshopper, inputs geometric curves and parameters to generate the geometry and attributes of elements. • Adaptive human-robot collaborative fabrication: Uses RGB-D cameras and laser displacement sensors for material detection and geometry reconstruction, with robots executing precise cutting and placement. • Integrated design-to-fabrication workflow: Maintains a feedback-driven production chain, continuously updating the computational model to ensure fabrication accuracy.

The experimental design involves implementing the Timbrelyn case study using a combination of reclaimed and new timber. In the experiments, the robotic system uses RGB-D cameras and laser displacement sensors for material detection and geometry reconstruction, executing precise cutting and placement. Results show that the Timbrelyn case study demonstrated 1,838 timber elements, with 91.1% robotically cut and placed, showcasing efficient material reuse and precise fabrication capabilities.

Experimental results show that the Timbrelyn case study demonstrated 1,838 timber elements, with 91.1% robotically cut and placed, showcasing efficient material reuse and precise fabrication capabilities. By combining reclaimed and new timber, the framework successfully addressed material inventory uncertainties, achieving efficient material utilization. The adaptive co-robotic assembly method reduced on-site construction complexity by shifting labor-intensive tasks to off-site prefabrication.

The framework can be directly applied in architectural design and construction, particularly in material reuse and low-carbon building. By combining reclaimed and new timber, the method effectively reduces raw material demand, lowers carbon emissions, and extends the lifespan of bio-based products. The framework can also be applied to the reuse of other material types, promoting the construction industry's shift towards circular economy models.

Despite the significant advantages of the framework in material reuse, it may face time and cost constraints when dealing with large-scale material inventories. Additionally, robots may struggle to handle overly complex or irregular timber shapes, and the framework's adaptability may be limited by the capabilities of robotic hardware and software. Future research directions include further optimizing the framework to handle larger-scale material inventories, developing smarter robotic systems to handle more complex material shapes, and exploring the reuse possibilities of other material types.