Published Paper Details:

The rise of autonomous and semi-autonomous machinery in industrial settings necessitates advanced safety mechanisms to ensure smooth operation while preventing collisions and protecting nearby workers [1]. This project proposes an AI-based Extreme-Edge Collision-Avoidance System utilizing a Temporal Convolutional Network (TCN) deployed on a Raspberry Pi microcontroller. The system integrates multiple sensing technologies, including LIDAR, Camera Modules, Ultrasonic Sensors, and Infrared Sensors, to comprehensively monitor the machine's surroundings [2]. At the system's core, the STM32 microcontroller processes real-time data from the sensors via a driver circuit, ensuring ultra-low-latency response. AI-based monitoring runs on a Raspberry Pi, analyzing time-series sensor data using the TCN model to detect potential hazards in real-time [3]. The AI algorithm predicts collision risks and enhances situational awareness, ensuring timely interventions critical for industrial safety [4]. Upon detecting an obstacle, the Raspberry Pi triggers immediate corrective actions, controlling machinery operations while engaging an alarm and alert system to notify nearby workers. The integration of Raspberry Pi extends computational flexibility, supporting data logging, visualization, and remote monitoring, while enabling machine learning model updates [4]. The system ensures robust performance in noisy and dynamic industrial environments by leveraging sensor fusion and AI. Extreme-edge processing minimizes latency, optimizes energy consumption, and maintains a compact memory footprint. Designed for seamless integration into industrial applications, this real-time collision avoidance system significantly enhances workplace safety, reduces accidents, and improves operational efficiency [5].