Paper Title
Real-Time Cargo Tracking and Tamper Detection Using Edge AI

Abstract
As the global platform, the supply chain become more complex, real-time tracking of cargo and tamper detection become essential for the security and integrity of the shipment. Usually, centralised cloud-based monitoring systems are employed, which are costly, cause delay, and require dependable network connectivity. An Edge AI-based cargo tracking and tamper detection system is presented here to address these challenges by using low-power IoT sensors, edge computing, and deep learning models to give real-time alerts and analytics. The suggested system combines GPS, accelerometer, temperature, humidity, and tamper sensors placed on cargo containers. Edge AI hardware processes sensor readings locally through a light deep learning model to identify anomalies, e.g., unauthorized entry, shock, or environmental excursions. The system uses an event-driven architecture where only meaningful alerts are sent to a cloud dashboard, minimizing bandwidth usage and facilitating lowlatency decision-making. Experimental tests were performed in a simulated logistics test environment that replicated different realworld cargo transport scenarios. The results show that the system attains a 96.4% accuracy rate in tamper detection, reducing cloud dependency by 83% against traditional solutions. In addition, the response time in real-time was optimized by 58%, allowing for prompt interventions in the event of security violations. The results show that the inclusion of Edge AI into cargo tracking improves security, efficacy, and real-time monitoring aspects, thus emerging as a sound solution for present-day supply chains. The ongoing work will also include the system extension to multiple transport modes as well as implementing blockchain for robust data integrity. Keywords - Edge AI, Cargo Tracking, Tamper Detection, IoT, Real-Time Monitoring, Supply Chain Security.