Enhancing Safe and Precise Maneuvering in Autonomous Electric Vehicle

This research deals with the design of an autonomous vehicle navigation system based on a master-slave computational paradigm, combining Raspberry Pi 4 and Arduino UNO for visual perception in real-time, decision-making, and actuation. The Raspberry Pi 4, supported by a Pi Camera module, performs lane detection and near-object identification through image processing techniques, and then sends over extracted data through serial communication to Arduino UNO. Contrary to traditional autonomous driving systems that are based largely on monolithic processing architectures, this two-part control paradigm maximally splits computational loads, thus maximizing real-time responsiveness and efficiency of computation. The approach coordinates an optimal interaction between image-based navigation and multi-sensor fusion to guarantee improved trajectory planning and obstacle avoidance. In addition, the servo-actuated ultrasonic module provides an extended spatial detection range over traditional static-sensor deployments, thus allowing for better environmental adaptability, especially in small or dynamically changing environments.