Monitoring and Evaluating ECU Communication Data for Vulnerabilities Detection

In today's automotive industry, Electronic Control Units (ECUs) control a large number of functions in vehicles. Increasing cyberattacks on ECUs pose a danger to vehicle safety and security due to the rising dependence on automation and Electronic Control Units (ECUs). This study proposes a new way to find ECU attacks by using Python's large library and tools for automotive security. The three main parts of the approach are collecting and processing data, and finding outliers. To gather data, an Arduino and a CAN bus shield are used to make a CAN network and record the data that the ECUs send and receive. Among the preprocessing tasks is the determination of the time intervals for message reception among nodes and the standardization of data to account for variances caused by hardware constraints. Examining the slope of data plots is a technique used for anomaly detection. The compromised Electronic Control Unit (ECU) is detected by a noticeable change in the slope, which indicates a potential attack. A proof-of-concept illustrates the strategy's efficacy. By inserting an adversarial Electronic Control Unit (ECU), which interferes with normal ECU communication patterns, we can simulate real-life attack scenarios. This Python-based method finds different threats successfully and with high accuracy. This study makes a big contribution to the fields of cyberphysical systems and automotive cybersecurity by showing a reliable way to detect attacks on Electronic Control Units (ECUs) that using Python.