How Aptiv’s SVA Reworks Vehicle Electronics for Software-Defined Cars
In general, the rise of the number of features led to the increase in the number of electronics control units (ECU), wiring and software. However, Aptiv’s proposed architecture is very different, as it attempts to unite dozens of ECUs in several central units and move some functionality to the software running on a powerful computer.
There is a clear explanation why this approach was chosen. In today’s automotive architecture, there are about 70 small control units scattered around the vehicle and connected with the wire harness. The key advantage of the Smart Vehicle Architecture (SVA) is the fact that the same amount of functionality could be provided with several central control units placed next to the backbone of the wire harness. However, the goal of this architecture is not only improved packaging, but creating room for software-centric functionality.
Specifically, the architecture should allow reaching several goals: cost reduction, software-defined features support and vehicle variants management simplification. Due to smaller number of control units and lightweight wire harness, the new architecture would help to save the costs associated with material and vehicle assembly. Moreover, it is especially important, because wire harness is heavy, costly and complicated to optimize in case of the vehicle with numerous features added over the years.
Moreover, the new architecture would provide zonal controllers, which would make it possible to connect sensors and actuators in an alternative way. While in the current system devices were connected to particular ECU, in the future they would be connected to the zonal controller, which, in its turn, would connect them to the central computer via high-speed communication channel. Aptiv calls this concept “the separation of I/O from compute”. As it can be seen from the engineering perspective, in this case the central computer would carry out software-related tasks, while zonal controllers would connect the physical world to the computer.
Separation of tasks is the reason behind tight connection of centralized architecture development and software defined vehicle development. With decreased dependence on particular hardware, it would become much easier for automakers to use software for providing updates.
Specifically, Aptiv sees several areas of application of SVA: over-the-air updates, advanced driver assistance, infotainment and energy management. The key idea of this concept is the fact that instead of vehicle electronics as a set of separate functional blocks, the vehicle would be much easier to update after its production.
Surely, there is an issue of maintenance and fleet. Consistent architecture in several models will simplify the diagnostics, especially for fleet operators and mobility services with large fleets of vehicles standardized electronics architecture will standardize the process. Though it would not simplify the complexity of the service itself, but it would transform it from module-to-module diagnostics to software/networking.
The key promise of Aptiv is the fact that the SVA will help to solve the whole range of issues from wire harness design to computing/safety electronics. This point is crucial, as far as transition to centralized architecture is not only computing problem, but packaging, manufacturing and validation problem. Transition to zonal architecture affects harness routing, controller placement, thermal management, software integration and functions partitioning.
The key thing is that the transition is not complete redesign for each automaker. Aptiv sees SVA as evolving blueprint, but not ready solution. This promise is realistic, as far as most manufacturers would not be able to make the shift to centralized electronics architecture in one development cycle new architecture would have to coexist with the legacy systems, supplier relationships and engineering strategy of the brand. Practically, it means that the implementation of each architecture would be compromise between target architecture and platform.
That is why Aptiv is not the only participant of this market. The concepts of Bosch, Continental solutions and in-house OEM architectures are very similar they are based on the concept of fewer control units, central computing, zonal aggregation and better support of software updates. In this market, the competition is not about problem recognition, but about implementation of the architecture in production without causing additional problems related to validation, cybersecurity or maintenance.
SVA is presented by Aptiv as “brain and nervous system” of the vehicle. Of course, it is marketing language, but the systems approach is right. SVA combines compute, connectivity and electrical distribution, instead of separating those layers. For US automakers and suppliers, the shift to the centralized vehicle electronics architecture becomes more and more of the platform engineering decision, rather than features story it has implications for assembly, diagnostics, update strategy and model scalability.
The key thing is that the shift to zonal and centralized electronics architecture is not only additional software introduction, but hardware sprawl elimination, which complicates the software implementation. If SVA concept turns out to be true, then the value of the approach is not in particular feature. It is the vehicle architecture, which becomes easier to manufacture, update and does not get loaded with wiring and control units of the previous generation of the automotive electronics.
By Robert McKinney — Editor-in-Chief for AMI’s automotive and mobility coverage, with a mechanical engineering background and a decade reporting on powertrain systems, EV innovation, and global vehicle manufacturing.
