Image: redsharknews.com

Central operating systems – what do they mean for driverless cars?

Technology and Business

Phil Brown

Phil Brown

4-05-2020

       

If you were to travel back to the 1950s, when the world’s first true (it looked nothing like a modern-day OS, by the way) was helping staff to crunch numbers at British food and catering company J. Lyons and Company, you would have been laughed out of the room if you said that one day cars would need a way more complex operating system just to get it from A to B.

Of course, operating systems have been present in  cars for many years now, from the menus on the first digital stereos to the built-in in-car entertainment and satellite navigation systems that are offered as standard on almost every new car these days. However, these operating systems simply aren’t future-proofed, and they don’t manage the actual operation of the car itself  – which we’ll get onto later.

 

Although there are already joint approaches between three (and more) of Germany’s biggest automotive manufacturers to try and catch up with Tesla, there that BMW, Daimler and VW are working on a centralised operating system for driverless cars. So why is a collaborative operating system so important to the trio?

 

 

 

The upcoming driverless car revolution

In the next decade, there are two huge changes that automotive manufacturers face: the electrification of vehicles and the next level of autonomous driving that sees our control reduced either completely, or significantly.

 

The system currently responsible for controlling acceleration, braking, steering and avoiding obstacles, as well turning up the heat or turning off the radio is of course the human. Although we’ve become pretty adept at driving, we still bring several natural failures to the wheel, including a sudden inability to stay awake on long and boring journeys, or that we simply can’t react quickly enough to the person in front changing lanes at the last moment.

 

A central operating system would need to take our complexities, change them for the better, and apply them 100% accurately every time. For example, humans accelerate in several different ways. Sometimes, we ease ourselves up to the speed limit, not really noticing the effect that slower acceleration has on our fuel tank. Other times, the music is turned up and we put the pedal to the floor, enjoying the G-forces as we fly down the motorway.

 

An operating system, however, will take a very different approach to our base enjoyments or rudimentary calculations to save fuel or enjoy the open road. At the forefront, safety will be the main objective (if we’re handing control over to an operating system, it needs to be better than us), so don’t expect to be pressed into your seat with the wind in your hair. Steadier acceleration is not just safer, but more efficient too, plus it’ll make braking easier if something does get in the way (another factor that will need to be managed by an operating system).

 

Aside from the actual driving element, there are plenty of other things that a driverless operating system would need to manage. When the battery runs flat, or our super-futuristic fuel cell is out of juice, the operating system will need to calculate the remaining journey, and maybe shift around some power usage to get us to our destination without charging (think of the ‘low power mode’ you may have on your smartphone).

 

We’re used to our cars telling us to top up the air in our tyres or visit a mechanic for an oil change, but what happens if the car suddenly becomes responsible for getting those jobs done? Would cars divert directly to the garage in the interest of safety, making us late for work even though the maintenance isn’t particularly vital, or will be have the opportunity to tell the car to wait until the journey home, like we often  do now (note: we hate to sound like a parent, but please never do this - always rectify issues with your car as soon as the alert appears on your dashboard). 

 

And what about if manufacturers give us the option to take control of the vehicles, maybe on roads that haven’t been ‘fully mapped’ yet? Will the operating system restrict the throttle in an attempt to keep us safe, or twist the wheel in our hands if we head towards a tree? If level five autonomy, where humans are taken completely out of the control equation, then it becomes less about how the driver gets the car from A to B and more about how they spend that additional time as the operating system does that for them.

 

As we step away from gear sticks and steering wheels, we also step towards , the way we interact with the vehicles we travel in will also change completely. This is already happening in high-end cars, with internet connectivity, TV screens in headrests and the kind of mod-cons you’d expect to see in a living room all now very common.

 

To understand just how critical operating systems will become in driverless vehicles, we need to think about how we’ll use vehicles differently. Currently, we may input a destination into a smartphone or satnav, but this could quickly turn into predictive destination choices, or even a simple voice command to tell the operating system where you need to go.

 

Then there’s in-car entertainment. Rather than entertainment, we could instead watch something, work, or use a screen rather than look at the road ahead.

 

So why the collaboration? Wouldn’t it be more advantageous for each manufacturer to develop their own operating system that would answer these new needs and create opportunities to take a bigger chunk of the ‘value cake’?

interior driverless vehicle
Image: m.futurecar.com

The challenges

One of the biggest reasons for car manufacturers looking to collaborate is the cost. Automotive manufacturers are great at creating vehicles but often lack the expertise needed in the fields of operating system development, UX and UI design and the actual implementation of highly complex software that only gets more complex as we electrify and automate. The upcoming reliance on operating systems is one of the reasons why many manufacturers to harness and attract the skillsets that are in demand in multiple industries.

 

By splitting the costs and workload of developing the ultimate automotive operating system three ways, the manufacturers can not only save money  but also break down the barriers between themselves and the competition to create interoperability that their customers will appreciate . The  ideas, innovation and existing knowledge that exists within three of the world’s biggest carmakers is certainly formidable, so it becomes an attractive proposition to work together.

 

But the bigger driving factor is time. Developing a whole new operating system is an unfathomably big job, especially if it is going to be responsible for the safety of passengers. Take Apple’s iOS, a very simple operating system that drives around 27% of the world’s mobile devices. Although it has been around for more than a decade, each iteration requires thousands of hours of work and development, and that’s just for a handheld device. Scale that up to something that will operate entire vehicle, including keeping passengers safe and hosting the software that will actually drive it, and the scope of the work required becomes clear. The problem here is also that some of the ‘traditional’ manufacturers are behind. In a big way.

 

As the driving world makes its way towards level five automation, a few names have already made huge strides in terms of usability and customer interaction. The most commonly known, Tesla, already has smart driver-assist vehicles on the road, with their own Linux-based operating systems more than capable of providing an excellent in-car experience. Then there are the tech firms and software development giants who are also seeing the advantages of entering the automotive market.

 

Take graphics and computing company Nvidia. Famous for microchips, GPUs and graphical interfaces rather than the commute to work, the US tech firm is getting to grips with not just operating systems, but also the software behind the driverless vehicle itself with its Nvidia Drive platform. The software is designed to go beyond just a UI and some fancy bells and whistles, but to immerse the passenger fully with the journey itself.

 

This is where BMW, VW and Daimler are likely concerned. Despite having exceptional expertise in controls, handling, materials and everything else that goes with driving, they cannot compete with the level of AI, leveraging of processing power, software creation and access to next-gen technologies that are nothing new to a company like Nvidia.  The next big question mark for car manufacturers is whether the collaborations go even further, with more cooperation between tech companies and car giants. We’re already seeing driverless  technologies being explored and developed by the like of Waymo (via Google), Apple’s unrealised  Project Titan and Russia’s answer to Google, Yandex, so we should be asking when carmakers will extend an olive branch to tech firms get in on this extensive innovation.

 

 

 

What driverless operating systems mean to the consumer

Choosing a new car comes with a huge range of choices. After the ‘hardware’ choices have been made, including fuel type, engine capacity, make or model and vehicle type, the customisation and configuration options can be endless. But with a driverless car, the options may be a lot more ‘superficial’ than you may expect.

 

With a computer at the wheel, all aspects of performance are likely to be removed, especially if an efficient battery array replaces the engine (which is a given, at least until fuel cell is ready for mass production). This means the look and feel of the car will certainly outweigh any decisions made based on 0-100mk/h times or handling abilities. The magic button on the dashboard that turns the economic family vehicle into something sportier at the weekend becomes pointless if a super-smart operating system is trying to get the maximum efficiency out of a single battery charge.

 

These changes pose quite a large question for consumers. Many of us opt to pay extra for vehicles with a little extra performance, or indeed an engine with a larger capacity for when we want to enjoy faster speeds. If the car is both electrified and controlled by an operating system, then certain cars’ USPs are suddenly eroded thanks to the ability to go from ‘safe and slow’ to ‘sporty and fun’ by tapping a button, rather than upgrading to a whole new car . With often huge price differences between basic models and top-of-the-range performance vehicles, we could see a new middle ground for both pricing and features that requires manufacturers to think of new ways to differentiate themselves from the competition.

 

Then there’s the interior, which will be where sales are made and lost. With no requirement to look at the road ahead, drivers (and passengers) will be able to do a whole host of different activities from reading and scrolling to playing games and watching movies or TV. Indeed, the car could become a living room, an office or somewhere to sleep, with all of the accompanying tech that goes into these rooms at home suddenly desired in the car.

 

This is where a common operating system starts to make a lot of sense. If you’re an Apple user, then borrowing someone’s Android device can be challenging, even just at the lock screen, but after a few minutes it usually becomes clear. The problem amplifies when we go from adjusting the timer on the electric oven to configuring the settings on our new gaming console.

 

By centralising, using and customising car operating systems goes along the mobile route – following the user’s actions and intuition to deliver usability that can be guessed, rather than learned. The benefit to the user is that they’ll be able to mirror their setup in their own car, in a taxi or in the hire car when they travel to a different location. But again, the things that set certain vehicles apart from the competition become ‘vanilla’, leaving gaps for designers to fill to make the journey ‘different’ for consumers willing to pay more.

 

The bigger picture, however, also requires us to look not just at our cars, but the road as a much larger ecosystem. If our car is responsible for navigating across a busy intersection for example, it will need to communicate with the other cars, heavy goods vehicles, buses, flying taxis, drones and of course the humble pedestrian. This is where the operating system gets another extra-complicated layer; a much bigger system in which it needs to slot in.

 

If there were 50 different operating systems all linking together to try and create a safe and efficient traffic system, then the problems already become apparent (try linking your Apple device to your Samsung television). With a centralised system, communication becomes a lot easier, especially if there’s a central hub that is responsible for tracking vehicles and making sure they’re not on a path that will result in an accident, or an inefficient journey.

UK Auto drive artificial intelligence
Image: Wired.co.uk

Will we need cars at all?

A huge question that will certainly dictate approaches in the very near future is the need for individual cars at all – or indeed the need for a personal vehicle whatsoever. With all of the design features, performance aspects and ‘heritage’ stripped out to make way for comfortably cabins and tech distractions, the choices are narrowed enough to the point where a ‘one size fits all’ vehicle makes more sense.

 

And with nose-diving running costs, manufacturing costs and of course the lack of driver needed, the cars we take to the office every day or into town at the weekend could become even more centralised. For the 75 million Uber users out there, the vast majority of the taxi fare goes towards the driver’s wages.  Take this away, and it suddenly becomes more economically viable to simply take a driverless taxi everywhere than buy and expensive vehicle outright (at the cost of having to find new jobs for millions of taxi drivers across the planet). Scale this up, and we could get to the point where  . With vehicle markets changing from personal ownership to a lift-sharing or ‘pay per ride’ model, car manufacturers are faced with a whole new set of challenges on top of and beyond the new tech and research that goes into something like a shared operating system . 

 

 

Would you welcome a common operating system and centralised form or car ownership? Or does the idea of losing the elements that make cars so special fill you with dread? Join the discussion at 2025AD, the internet’s home for autonomous driving opinion.

Engage

<div id="hs_cos_wrapper_Engage_" class="hs_cos_wrapper hs_cos_wrapper_widget hs_cos_wrapper_type_inline_text" style="" data-hs-cos-general-type="widget" data-hs-cos-type="inline_text" data-hs-cos-field="submit_your_story.icon_text">Letter</div>
Submit your story

Become part of our autonomous revolution and submit your stories, images and videos

Submit your story
Megaphone
Stay informed

Stay up to speed with our weekly briefing. Enjoy autonomous driving content direct to your inbox

Join our weekly briefing
Connect
Connect with us

Follow us on our social networks for up to date information and thoughts on automated driving