Take a leap: virtualization in future development

The automotive industry is facing various challenges, from complexities of bringing software-defined vehicles on the road to increasing technical capabilities for autonomous vehicles. Due to the processing power needed to run software-defined functions, software creation goes beyond traditional methods and often, extensively uses artificial intelligence (AI) and evolving sensor technologies. Additionally, serverbased vehicle network architecture and service-oriented base software might cause a shift of sourcing strategy to software only. This is where virtualization becomes part of the equation and leads towards a possible new vehicle lifecycle model that challenges our current understanding across car makers, suppliers, and their business models.

To gain acceptance from drivers and passengers and the ever-stringent authorities, we must demonstrate all this new technology is reliable and safe. How do we decide the completed testing and validation is enough to prove our vehicles are safe and ready for mass production? What is the best balance for quality, cost, and time for a vehicle launch? Virtualization may become mandatory in future architectures to redesign the vehicle electronics networks with virtual resources ranging from processors to operating systems to data servers.

Virtualization must be managed carefully to ensure the benefits outweigh the additional risks for vehicle development. There are many aspects of virtualization that need to be considered from para to full virtualization or critical to non-critical software aspect separation. It’s time for the automotive industry to challenge its value and embrace virtualization in some form.