Membership Function Optimization in Fuzzy Logic Controllers for the Performance Enhancement of ESP32-Based Line-Following Robots: A Systematic Review and Performance Analysis

Abstract

The rapid evolution of autonomous robotics highlights the persistent challenge of controlling non-linear dynamics in line follower robots facing environmental disturbances. Traditional linear controllers exhibit inherent weaknesses, prompting a paradigm shift toward Fuzzy Logic Controllers (FLC). However, conventional FLCs predominantly rely on heuristic tuning for membership functions, resulting in suboptimal parameters and overtraining. Furthermore, advanced optimizations frequently ignore the hardware constraints of edge computing devices like the ESP32 microcontroller. This study presents a systematic literature review analyzing the optimization of fuzzy membership functions to enhance ESP32-based line follower performance. Utilizing the Theory of the Consolidated Meta-analytic Approach (TEMAC), this research systematically evaluates empirical studies from 2022 to 2026. The findings reveal that triangular membership functions are the most theoretically and computationally optimal geometry for the ESP32's Arithmetic Logic Unit, minimizing the processing latency without sacrificing the kinematic responsiveness. Moreover, the integration of population-based metaheuristic algorithms, specifically Particle Swarm Optimization (PSO), revolutionizes parameter tuning by systematically eliminating human bias. PSO-optimized FLCs significantly reduce the Root Mean Square Error (RMSE), virtually eliminate actuator overshoot, and drastically accelerate settling time compared to manual calibration. The implementation success ultimately depends on effectively managing the ESP32's dual-core Real-Time Operating System and integrating the ESP32-CAM visual sensors for precise centroid extraction. The review concludes that future automated guided vehicle designs must deeply synergize PSO-optimized fuzzy parameters with the computational limitations of silicon edge computing. This strategic hybridization produces highly resilient, power-efficient, and agile robotic prototypes capable of robust navigation across dynamic and highly unpredictable industrial environments.