The aim of this study is to synthetically generate IoT digital supply chain network traffic and develop the concept of an intrusion detection system. Consistent technological innovations brought on by the digital revolution have resulted in a practically infinite variety of cyber threats, making IoT security a continuously challenging issue. Traditional system security solutions can't ensure the same level of security within IoT devices as they are often agent-based and their architecture differs. In order to achieve the goal of the study, challenges of network data generation were examined. A supply chain-specific IoT environment was modelled, network traffic was extracted, and different machine learning models were tested. IoT devices employ unusual traffic patterns, but by applying contemporary classification techniques to it, it was possible to create machine learning classifiers that can detect up to 99.99% of intrusions. It also showed reliable results even when tested against unpredictable behavior from different cyberattacks that were introduced into the network.