Today’s vehicles are becoming more and more computer and less and less automobile. Software-defined vehicles (SDV) shift a lot of the hardware in a vehicle, including relays, ECUs, and switches, to software-based platforms. This, theoretically, allows automakers to make changes and improvements to the vehicle after it’s been built and even after it’s been sold.
This shift to computer-based everything is happening despite the problems with sourcing chipsets and other computer hardware. What automakers have been doing, in fact, is shrinking the number of computers in the vehicle via consolidation while at the same time increasing the amount of processing that the vehicle does.
In effect, this means that SDV is a centralization of computing power in the vehicle. Using more powerful processors instead of smaller, cheaper, control units. This supposedly simplifies the architecture and reduces complexity. It also leads to some big concerns.
I’ve often complained about the move from intuitive buttons and switches on the dashboard to big screens with multi-layered menus. When that big screen decides not to work, half the vehicle’s functionality goes with it. What happens when the lone chip running the whole car decides it doesn’t work today? Or gets hacked?
Worse.. what happens when that SDV becomes subscription-based? Which we all know is inevitable as this progresses. I talked about VaaS earlier. Even ownership won’t necessarily guarantee access. Think about how often your current subscriptions decide to not work or log you out. So that you’re required to use that stupid TV remote to “type” in a lengthy (thanks to requirements) password just so you can see if Netflix has anything worth watching.
Now imagine that’s your car. Good times.
It’s bad enough that a lot of vehicles today take over a minute to recognize any Bluetooth device. Wait until that’s the minimum wait just to start the vehicle and drive somewhere. The enshitification is coming.
And all of these systems will be connected to the world around them. A key component of self-driving and semi-autonomous vehicles is that ability to connect. And a big part of SDV is being able to do over-the-air (OTA) updates. Not to mention the data gathering and dissemination that will inherently grow as software defines more and more of the vehicle. Probably so you can be served ads delivered by that OTA.
Lets not even start into the whole “machine learning” Skynet thing here.
Some upsides to SDV are also obvious, though. I’d be remiss to ignore those. Vehicles will, hopefully, hold higher resale value because they’ll remain technologically current for a longer lifespan. They’ll allow more customization for the consumer as well. They’ll also, it’s opined, make developing and producing vehicles a lower-cost prospect as the software and much of its hardware becomes platform-agnostic. Those gung-ho about SDV are predicting that it will reshape automotive in the same way that smartphones changed consumer electronics.
On that last point, I’m less enthusiastic. A car is, in the end, still a car. Some wheels, a way to steer it, some kind of propulsion, and a place to sit. That’s the basic definition and none of that is electronic. Hardware will still matter. Computers can’t make suspension happen. Nor can they change steering and movement dynamics. Or comfort systems. Or energy storage. Or make the whole car look cool or feel fun. Most of that is mechanical hardware.
The upside for SDV enthusiasts is that the majority of vehicle buyers and owners think of their vehicle as another appliance. It vaguely defines their personality while they’re near it, but it mostly just sits around until they need to get from here to there.
However it turns out, SDV is already happening and it’s probably going to be a fairly rocky road as it does. And for my fellow curmudgeons who say that they’ll just keep driving their 1990s whatever.. Well, those aren’t gonna last forever either. Sucks, I know.
This article originally published on the AaronOnAutos Substack.
