Cut Sequencing Algorithm for Safely Disassembling Large Structures

Disassembly and fragmentation are key operations in the dismantling and recycling of decommissioned structures such as aircraft, vessels, and buildings. Often, such operations are hazardous requiring careful planning for safe execution based on the experience and intuition of workers and forepersons. We propose and devise an algorithm for the automated sequencing of cuts to disassemble large structures. Using feedback from physics-based simulations and a mathematical model for safety, our algorithm performs sequential decision-making yielding the order of the cuts on the structure and the corresponding safe standing positions of the cutter (representing a worker or a robot). Our goal is to determine a sequence of cuts and cutter locations to maximize safety for the cutter and the environment. We establish the optimal solution via exhaustive searching, and design a greedy decision scheme to reduce the search runtime. Using our evaluations in simulation, we compare our greedy decision scheme against exhaustive searching and random searching, concluding that it satisfices the goal with high safety scores and low runtime.