Conferences and supporting programme
On the Road Towards Autonomous Driving – Simplifying AI with Virtualization Technologies
Incredible artificial intelligence techniques and advanced communication patterns will be at the heart of autonomous driving systems. The computational requirements, as well as the software complexity and related integration, are the main challenges facing the automotive industry. Silicon manufacturers have released powerful clusters of processor cores envisioning hardware consolidation. This new approach opens complete new applications, services and business models. Separation and Virtualization for embedded systems is the key enabling technology in addressing these challenges. Hypervisors are the underlying component of the virtualization technology, being responsible for the scheduling of the heterogeneous virtual machines (VMs) with different real-time constraints (e.g. AUTOSAR, Linux, Android). In general, hypervisors provide strong separation, ensuring freedom from interference within mixed-criticality systems. Virtualization also enables secure partial software update, acting as an extra layer of security beyond software platform-specific hardening. Furthermore, virtualization can shield a highly connected vehicle, enhancing the mitigation of cyber-security risks. Especially combinations are interesting, if during high volume data processing, for example, intrusion detection and at the same time AI components can simultaneously use the data stream transported once. Significant simplifications and performance improvements are possible.
--- Date: 27.02.2019 Time: 3:00 PM - 3:30 PM Location: Conference Counter NCC Ost