Connectivity steers toward the software-defined car
Automotive industry stakeholders are heavily engaged in the creation, refinement and application of connected/autonomous-vehicle (CAV) technology in the quest to develop next-generation vehicles.
According to James Hunt, CEO at aicas, advances in solutions that enable over-the-air (OTA) programming for vehicles are already answering stakeholder concerns relating to revenue for carmakers, third-party ability to introduce features and services and planning for future regulatory compliance. However, although OTA solutions are changing the outlook for system updates, upgrades and bug fixes of software assets, he believes that widespread adoption will be fueled by advances like the rollout of 5G networks capable of reliably handling massive amounts of data, particularly when accessing a complex system already running multiple, critical functions simultaneously, and which may be travelling at more than 60mph.
Moreover, as OTA advances open new connections to vehicles, he stresses that the management of cyber security and identity and access authentication systems becomes ever more important. He said: “To keep ahead of the asymmetrical threat hackers pose, a system needs to be able to manage resources, know their identity and version, and be able to update them at a granularity far below today’s flashing of complete ECUs.
“A critical success factor for software architecture is going to be the granularity of these components. In today’s setup, most are monolithic functions that require a full system exchange that can take from several minutes up to several hours or even days. A microservice-based approach not only provide more flexibility of the overall software architecture, but also enables small updates within a few minutes on all levels of the car’s software.”
Complexity and risk
Elsewhere, Pedro Lopez Estepa, senior market development manager, automotive at RTI, observes that the development of autonomous vehicle connectivity software is inherently challenging and risky for a number of reasons. To begin with, since they are built from many subsystems, vehicles require full interoperability between components and, in order to provide consistently optimal performances, data must also flow correctly, reliably and with extremely low latency.
To read the complete article, visit TU-Automotive.