Securing the Message Queuing Telemetry Transport protocol (MQTT) in Industry 4.0 is a theoretical concern and a practical necessity. It is crucial for protecting sensitive data, ensuring real-time communication integrity, maintaining operational continuity, and preventing unauthorized access. It mitigates Cybersecurity threats, meets industry compliance standards, and safeguards interconnected industrial systems from disruptions and attacks. This paper introduces a novel approach to studying threats in IoT communications. Our method involves the generation of synthetic traffic of MQTT packages using a sophisticated Backend and Frontend software architecture. By harnessing the power of artificial intelligence, we can gain valuable insights into the security of IoT devices and brokers. Our method uses a visual platform to virtually recreate typical industrial IoT sensors that measure temperature, humidity, pressure, and speed, among other variables, with changes in the resolution and number of packages per second generated. The MQTT traffic is pointed to an IoT Broker and automatically intercepted by a virtual Sniffer that stores package behavior in a parallel dataset. Then, several classification algorithms are trained to identify whether an attack like a Denied-of-Service (DoS) is present in real-time. Experimental results show that our synthetic MQTT traffic effectively simulates various attacks on IoT communications, which can be accurately identified using classification algorithms, achieving high-performance detection.