Software-Defined Vehicles Should Be More Open, Modular, and Upgradable

Sign up for daily news updates from CleanTechnica on email. Or follow us on Google News!

Last month, I shared the story of how GM arrived at the decision to abandon Android Auto and Apple CarPlay for the next generation of EVs. In short, iffy reliability was a drag on safety, so GM’s engineers figured that bringing it all in-house would make for a safer and more reliable way to deliver apps to drivers.

I’ve been stewing on that for weeks now.

In some ways, GM’s absolutely right. I own a Bolt EUV and the experience with Android Auto has been less than stellar. Instead of being able to just get in the car and drive, I find that sometimes it just won’t connect wirelessly. So, I have to always leave a cable plugged in and be ready to have a passenger plug my phone in for me. This was especially painful on road trips.

But, at the same time, my Bolt’s software sucks. It’s very basic, it doesn’t offer superior navigation experiences like Google Maps or Waze, and it offers no way to add other data sources (like ABRP EV charge planning or OBD-II data) into my situation awareness as a driver. So, GM had better get the company’s app ecosystem right if alternatives like Android Auto aren’t going to be available.

There are better experiences available, especially Tesla’s offerings. Instead of needing to rely on a phone to provide important things like trip planning and connectivity, Tesla has most of that well-figured out and well thought-out. But, even the best still has drawbacks, as some owners have learned when needing to upgrade or replace their main displays.

So, even though EVs are still fairly new, and my Bolt isn’t much more than a year old, we’re already seeing vehicle electronics grow obsolete rapidly, with phone connectivity being a major saving grace in some cases. In others (like the 2011 Nissan LEAF I used to own), they’re stuck feeling ancient and even stop connecting to cellular networks.

In other words, as EVs age, they’re going to be like your old smartphone that’s kicking around in that drawer of old electronics that you can’t bear to recycle yet. They’ll work, and get people from A to B  reliably for years to come, but their software and hardware will start to feel very dated.

This Doesn’t Only Happen To Cars

While cars are a lot cheaper than military fighter jets, there are some big commonalities. As cars get older (again, I’ll use the 2011 LEAF I used to drive), it makes no economic sense to upgrade their systems to what the latest car has. There’s just no upgrade path. It turns out that this has even happened to $100 million planes.

Despite immense costs, the earliest versions of the U.S. military’s F-22 have been retired because it just didn’t make economic sense to upgrade them to the latest version. Why? Because defense contractors simply didn’t make any plans for future upgrades, thinking that the Cold War pace of fighter development would have put them onto the back burner or into the boneyard.

But, the Air Force has learned from that mistake (among others) with the F-35 program that had a lot of problems early on. To keep the planes relevant for decades to come designers thought ahead about at least some upgrades, and now the military’s investing money into making the plane even easier to upgrade. Why? So that the planes can be useful for decades into the future without giving them basically a full rebuild every few years.

For the next generation of planes, comprehensive and easy upgradability will be built-in from the beginning:

In short, instead of having to spend billions like the U.S. F-22 and Taiwanese F-16 programs did for upgrades, things like OTA updates, modular architecture, and over-built computing systems can keep the planes not only relevant, but cutting-edge for decades into the future.

This will give future planes an even better value for the massive amount of taxpayer funds spent.

What Automakers Should Learn From This

Sadly, most automakers have maybe a 5-7 year operational design life in mind for new cars. Designing them for a 15-20 year life (something many EVs will be able to achieve) means lower vehicle sales and a big hit to planned obsolescence. But, it would be far better for the environment to make cars not only more durable, but built with a plan to remain useful and desirable for a longer lifespan.

Beyond the environmental, there are some other key benefits to building vehicles with a more modular and upgradable electronic architecture.

First off, it allows greater customizability. If you want to add a better driver assist or self-driving system, a better infotainment system, radio gear (especially important for government buyers), sensors, or anything else, that should be a lot easier.

While we aren’t all trying to do Cannonball runs, the trouble and expense (approaching $100,000) Alex Roy had to go through to upgrade his BMW to make his record run across the country is instructive. We’re not all upgrading cars to be more like a fighter jet, with intelligence, electronic warfare, and night vision systems, but wouldn’t it be nice if you could just buy sensor modules and let the car’s already hefty computer read the sensor data?

Or, even better yet, why not have a main automotive and infotainment computer that can put it all together (sensor fusion) and control everything? This sounds extreme, but it really isn’t. Imagine if someone wanted their Waze software to be able to add in radar detector data and automatically share it with other drivers? Or what if an Uber driver wanted to let the car give updates on a food delivery without having to use a phone mount?

Modular software and hardware with room for future upgrades and customization opens up a lot of possibility for general usefulness, lifespan of the vehicle, and the lowering of overall associated carbon emissions. 

It would be a big mistake for the next generation of EVs to not be prepared to take full advantage of this concept and deliver on these benefits.

Featured image by Jennifer Sensiba.