Robotic modular construction machinery with self-configuration systems: architecture, algorithms and economic efficiency

Abstract

The construction industry is facing global challenges related to low labor productivity, a shortage of qualified personnel, and growing demands for environmental sustainability. Modular robotic systems with the possibility of self-configuration represent a promising solution to these problems. This article examines the architectural principles and algorithmic approaches to creating self-configurable modular robots for construction, as well as the economic aspects of their implementation. Based on a comprehensive analysis of existing prototypes and technologies, a classification of modular construction robot architectures has been developed, including chain, lattice, and hybrid types. The use of multilevel reconfiguration planning algorithms based on genetic algorithms, deep reinforcement learning and swarm intelligence methods is justified. Empirical studies on the modular system prototype have shown a 37,8% increase in productivity and a 24,3% reduction in costs when performing typical construction operations compared to traditional methods. An economic model has been developed demonstrating that the integration of self-configurable modular robots provides an average return on investment over 3,2 years and a 31,5% reduction in construction waste. The methodology proposed in the paper for designing and evaluating the effectiveness of self-configurable modular robots creates the basis for their widespread introduction into the construction industry, contributing to improving the efficiency, safety and sustainability of construction processes.