Cycladic dry stone technique transformed to computational design logic

Abstract

Dry-stone walls have been a defining feature of rural landscapes for centuries. Built without mortar, they rely on precise stone placement and interlocking techniques to achieve stability and durability. Found across the world, including in the Cycladic islands of Greece, these structures serve various purposes, from retaining walls and agricultural boundaries to providing habitats for species ranging from insects to humans. The artisanry required for dry-stone masonry has traditionally been passed down through generations, yet it is increasingly at risk due to economic shifts, changing labour priorities, and the decline of traditional construction practices. This research focuses on the dry-stone walls of the Cycladic island of Sífnos and explores how computational tools can document, analyse, and potentially support their planning and construction. To digitize the stones, both LiDAR and photogrammetry were evaluated, with photogrammetry ultimately selected as the primary scanning method. These digital models were then analysed in Grasshopper, a parametric design tool, where machine learning techniques were used to evaluate their size, shape, and functional attributes, such as prósopa (stone faces). Based on these attributes, stones could be categorised by their intendeduse and assessed for placement. A key aspect of this study was testing whether digital classification could aid in real-world dry-stone assembly. A case study was conducted,where digitally categorised stones were used in the manual construction of a dry-stone wall. While the results demonstrated the potential of supporting the assembly process with computational methods, challenges arose applying these digital categorisations to traditionalstone masonry, which relies heavily on irregular shapes and intuitive decision-making. Furthermore, environmental constraints – such as lighting variations affecting photogrammetric accuracy – posed additional challenges that require further refinement. Beyond possible technical contributions, this research highlights the ethical considerations of integrating digital tools into traditional crafts. While computational methods can assist in documentation and planning, they should be treated as a complement rather than a replacement for artisanal expertise. The displacement of traditional knowledge, particularly in culturally identifying crafts, remains a pressing concern. As such, this study raises the importance of mindful implementation, ensuring that technological advancements support, rather than overshadow, the knowledge embedded in traditional artisanry. By bridging traditional dry-stone masonry with computational design methodologies, this research lays the groundwork for possible applications in cultural heritage preservation, sustainable construction, and adaptive reuse of materials. Potential uses include the repurposing of irregular demolition waste using similar methods, as well as further development in digital-assisted dry-stone construction forreconstruction or conservation efforts.