Particle Filter-Based Localization Algorithm for Autonomous Robots in Smart Factories

This research aims to study and compare the performance of four particle filter-based localization algorithms, Monte Carlo Localization (MCL), Auxiliary Particle Filter (APF), Rao-Blackwellized Particle Filter(RBPF), and FastSLAM implemented in smart factories using autonomous robots. By conducting a variety of simulations as well as real-world experiments, the paper assesses algorithmic performance in terms of accuracy computational efficiency, and responsiveness to change. FastSLAM shows itself as the best model demonstrating 0.09 of MSE and 15 ms of processing time among others. RBPF also demonstrates notable accuracy (MSE: It also focuses on its factorized state representation, accuracy, and computational efficiency. MCL and APF, which reach excellent precision and adaptive capacities, however, display slightly more computational expenses. The study offers useful information for practitioners and researchers looking for perfect solutions to robust autonomous robot localization in dynamic industrial conditions, thereby contributing towards the development of smart factory technologies.