Software Development Support Tools

At a certain automotive company, CAN communication verification work had become a major challenge in the development of in-vehicle device software.
Verification of communication content integrity is performed manually, resulting in frequent verification omissions and human errors.Demand for “dedicated tools for specific sites” capable of handling complex communication specifications was growing at development sites.

Therefore, we developed a one-off CAN communication test automation tool fully customized to the customer’s development environment and communication specifications.We implemented the following features to directly approach challenges in the field:

• Building virtual communication nodes
• Creating scenarios based on software behavior
• Automatic collection and analysis of communication logs
• Automatic determination based on communication content specifications
• GUI-based operation screens designed for anyone to use
• Support for communication protocol extensions

In response to customer requests such as “We want to automate the process of checking communication logs” and “We want to output details when errors occur,” we built an environment that virtually recreates CAN communication.Virtual nodes now communicate as substitutes for ECUs, enabling frame transmission and reception according to specifications, as well as timing adjustments.We also enabled the configuration of failure scenarios such as error responses and communication disconnections, creating a flexible environment where developers can perform a wide range of verifications without requiring physical devices.

We developed a tool equipped with real-time analysis capabilities to comprehensively detect diverse communication patterns in CAN communication, including normal frames, delays, communication interruptions, and error frames.Communication logs are automatically collected and categorized, and based on predefined judgment logic, they instantly visualize the type of anomaly and the timing of its occurrence.Developers can build verification scenarios themselves, as the determination conditions can be flexibly set within the scenario.

Prior to implementation, all CAN communication log collection, verification, and judgment were performed manually, requiring the experience of skilled personnel for anomaly testing.Detecting communication timing discrepancies and error frames was also difficult, and there were verification omissions on occasions.
After implementation, verification conditions could easily be set within the scenario, and through the automatic collection of communication logs and real-time analysis, anomalies could be detected immediately.The verification results were automatically output in CSV format, enabling their use for sharing information with others.