Search for a command to run...
With the advent of the Industry 4.0, the proliferation of IoT devices and artificial intelligence (AI) has accelerated the adoption of smart technologies in industrial safety, including accident detection and proactive risk management. However, conventional camera-based collision management systems suffer from blind-spot issues and high computational demands, which limit their practicality and result in high latency, making real-time operation difficult. In addition, sensor-based systems are constrained by limited positioning accuracy, often leading to reduced reliability and the generation of unnecessary alerts. To address these limitations, we propose the Intelligent Industrial Safety System (IISS), a ultra-wideband (UWB) and AI based feedback framework designed for proactive collision prevention. The IISS is consist of three subsystems: an indoor localization subsystem utilizing UWB anchors and tags for accurate positioning, a collision prediction subsystem employing an AI-based trajectory prediction, and a user feedback subsystem that delivers multimodal alerts through wearable devices. In this study, we designed and presented a proof-of-concept implementation of a IISS capable of predicting collision arising in blind-spot scenarios. In addition, tests were conducted in a worksite based on the proposed system. The results indicated that the system can predict potential collision scenarios and provide workers with intuitive feedback through vibration, auditory alerts, and context-aware text messages.