Recent advancements in Internet of Things (IoT) technologies, such as Narrow Band-IoT (NB-IoT) and efficient on-the-edge tiny machine learning (tinyML) systems, offer numerous opportunities for developing smart monitoring systems. The transport sector, in particular, can benefit significantly from these innovations. Tracking and detecting dangerous events in real time, especially for valuable goods, remains a challenging task. This paper describes a smart shipping system that utilizes a Neural Network to classify accelerometer data streams and NB-IoT to transmit real-time events and reports to users. We present an IoT prototype featuring a smart camera and a System-in-Package (SiP) that integrates cellular connectivity, offering advanced shipping monitoring capabilities.