Intelligent Multi-Sensor Data Fusion Framework for Autonomous Engineering Systems

Abstract

Autonomous engineering systems require accurate perception, intelligent decision-making, and adaptive control to operate safely in dynamic environments. This paper proposes an AI-driven Intelligent Multi-Sensor Data Fusion Framework that integrates heterogeneous sensors with deep learning, probabilistic fusion, and reinforcement learning for enhanced environmental perception, autonomous reasoning, and real-time control. The framework consists of four layers: multi-sensor acquisition, intelligent preprocessing, AI-based fusion and reasoning, and autonomous execution. By employing adaptive confidence weighting and context-aware learning, it improves object detection, localization, fault diagnosis, decision reliability, and system robustness while addressing challenges such as sensor uncertainty, synchronization, failures, and cybersecurity. The scalable architecture is applicable to autonomous vehicles, robotics, smart manufacturing, infrastructure monitoring, and Industry 5.0 cyber-physical systems, with future research focusing on explainable AI, federated learning, digital twins, edge intelligence, and trustworthy autonomous decision-making.