As vehicles have become more interconnected and software-defined, attack surfaces have become bigger and increasingly complex. In recent years, various vehicle intrusion techniques have jumped from theoretical concerns to real-world financial liabilities for OEMs. In this environment, intrusion detection systems (IDS) are essential to a secure automotive platform.
Data-driven models are an effective way to configure an IDS for many complex vehicle architectures. However, the effectiveness of an IDS is disproportionately correlated to the quality of the data used to develop a model of vehicle network traffic.
In this free webinar, Garrett Motion’s Volkan Deveci, Joseph Antoon, and Ioannis Deligiannis explore the unique challenges of data-driven IDS configuration and discuss the present and future of automotive IDS.
Key topics and takeaways:
- Complex data ecosystems – investigate the intricacies of handling vast amounts of data from various sources
- Reducing false positives – understanding the disproportionate cost of false positive anomaly alerts and how to prevent them
- Vehicle variants – building robust models that support multiple configurations
- The potential and drawbacks of intelligent systems in automotive IDS