Abstract
The built environment faces challenges of limiting resource consumption and emissions while addressing growing development demand and low labour productivity problems. It requires disruptive approaches to facilitate a transition to the circular economy and the introduction of several technological enablers to overcome challenges, such as smart data acquisition, data fusion, and design for robotic construction and deconstruction. This paper introduces a digital tool for robotic deconstruction planning of monolithic reinforced concrete structures based on automated extraction of data from scan-to-BIM at a low level of detail. A script operating on scan-to-BIM defines element connectivity and geometrical and physical properties for a network graph model used as the basis for the metaheuristic ant colony optimisation implemented to support the scheduling of human, crane, forklift and robot collaboration tasks. The methodology is explained and tested in three case studies with different levels of complexity, i.e. two synthetic flat slab structures and one real scan-to-BIM input. Cutting speed, effective crane utilisation and high human labour cost are the main parameters influencing the optimal deconstruction plan.