Title is a bit overblown and unnecessarily negative. Why can't we just be happy about the future. 'Cos it doesn't attract readers? 'Cos even seasoned contributers to Forbes have to express their petty, pissy attitudes?
The title should be, Don't forget Volvo, Nissan, BMW and others are working on driverless cars too. The future will be awesome.
This might be the worst written article I have ever read. What the heck is this guys point?
I'll tell you why it's being done by Google. A computer driving a car is an artificial intelligence problem. Google is fundamentally an artificial intelligence company.
Car companies are not artificial intelligence companies. They're not even software companies - just look at antiquated the average stereo/navigation system is in even the latest, most expensive cars.
I'd definitely rather have Google do it and license it back to the car companies than have the car companies do it on their own.
The software is going to be solved in the next 5-10 years. The hard part after that is dropping the sensor price enough to make this work (a lot of the SICK lasers are $35k+ still). But 10-15 years from now, we are going to have self driving cars for sale, no question in my mind.
Driverless cars on roads where people's lives are at stake is a social-legal-regulatory issue and even when the software is solved these issues remain. Even when there are cars capable of driving unaided, humans will be legally responsible for them - so we need to factor that in too
I don't think the op was arguing that car companies don't produce software. He was arguing that they don't produce it well, because it's not a core competency for the company. Car company profits are tied to industrial design, manufacturing, and logistics, not software.
His point about GPS is an apt one. My brother has the best SUV Infinity makes and the GPS is so terrible, he uses his phone (android) instead.
Suppliers to the car makers, like those that produce anti-lock brakes, do good software. Car makers typically have Silicon Valley offices. And they typically buy components from companies rather than rely on themselves. None of that means Google will make and driverless car that sells in the way the original article claimed.
Car companies write software that is incidental to their primary purpose. Software companies write software as a primary purpose. This is the definition of a software company. A more interesting argument is whether a software company is more likely to produce a successful self-driving car, which I believe the parent comment demonstrated to be likely by pointing out that it's an AI problem, and Google does a lot more with AI than car companies.
Those are very different kinds of software. Everything other than the last one is a real time OS problem with very known inputs and outputs. Validation is critical but manageable. And even then, most of them are outsourced to Bosch and other suppliers.
Unfortunately, this article exposes its author's near total ignorance of the background and history of the topic, and nothing else.
Driverless car projects have been around for 30 years. Anyone following European R&D in this area can easily cast their minds back to the DRIVE project
That's awesome. Where were they in DARPA's 2 Grand Challenges and the Urban Challenge [1]? We can also cast our minds to ALV [2] and NAVLAB [3], which were state-of-the-art in 80s and 90s respectively, so what? They demonstrated that a computer can control the car in a strict sense -- keeping it on the road and maybe stopping for obstacles, but driving is a bit more complicated than that. The real challenge is in navigating from point A to point B which observing rules of the road and interacting with other traffic participants safely -- something that was not demonstrated until 2007. Google's autonomous car program grew directly out of CMU's and Stanford's Urban Challenge projects.
Volvo pulled off the more impressive feat of platooning the three cars behind a lorry for 200 kilometres on a busy Spanish motorway.
It's not a more impressive feat by anyone's standard. Following the leader is a much simpler problem than navigating traffic independently.
Forbes.com = sensationalist headlines designed to grab clicks. Even their financial coverage is far more sensationalist than WSJ or techcrunch (and that's saying something ...)
Premise: "Car companies have been toying with driverless cars for years, so the idea that Google will produce driverless cars end-to-end is without merit"
What I can't understand is who thinks Google is going to be building these cars. Has anyone claimed that, even the referenced Forbes article? I don't think so. As others suspect, Google wants to offer the brains to these car companies, and tie them into the world of other Google services (Android phones, browser integration, ad injection, media streaming, etc).
Look at it another way: Microsoft supplies much of the tech to Ford, Hyundai, and others for their in-car "infotainment" systems, that doesn't mean when they started the CAR.NET initiative in the '90s they were trying to take on Ford and GM.
A second, follow-up point: If car makers have worked so long on driverless technology and made so little progress, what is Google doing better? I've seen the driverless cars in action, they work very well (though I elected not to leap in front of one).
Google Maps. The idea that a car maker is better positioned than Google to pull this out is silly; it's mostly about understanding geographical/geospatial data best, not so much how to make cars; that's already pretty well researched and understood and Google can partner with cars makers in an Android like model if necessary.
it's mostly about understanding geographical/geospatial data best
If Google's driverless tech is more dependent on geographical/geospatial data than monitoring actual conditions, it's going to fail horribly. Roads can change a lot day-to-day from storms, accidents, construction, etc.
I recommend reading up on how they create the "geographical/geospatial data" before casting doubt. It's similar tech that's used for identifying road signs, land widths, exits, on-ramps, etc...
I think that covers the hardest bits. Maintaining safe distances, self parking, and obstacle avoidance were nearly solved problems before Google started the driver-less cars program.
That article doesn't address my concerns at all - namely roadway changes caused by factors like construction, storm washouts, etc - that haven't been integrated into the geographical/geospatial data.
In fact the article was not anti-driverless cars and says Volvo will introduce some driverless features in 2014. It simply disputes the idea that this market is Google's for the taking
I'm excited about the prospect of driverless cars but at this point, even a regular car design by Google would get my money in a heart beat. I imagine a nice Nexus 10 as the main console and extreme environment friendliness.
> You might then ask why in the USA is this being left to a search company?
Google is a technology company. I'd prefer a tech company was working on this problem.
Driverless cars involve so much more than just an algorithm. It is the hardware integration, safety and reliability, regulatory and compliance and the worldwide scalability of the approach.
Google may understand the algorithm but they have nothing to contribute elsewhere.
i have no idea why you're being downvoted. auto and avionics companies go through hell to test this stuff at a level that involves the hardware and how it integrates with the software. this just doesn't happen in most software companies.
this is much more than a software problem. i've heard of people verifying opcode behavior on avionics processors to determine the correct implementation of the hardware - when your software relies on it it's fair game.
i don't doubt that google could do this if they made it a priority. but it's a huge issue that encompasses more than a "software algorithm." if the collisions sensors degrade over time doesn't matter if the software works.
I would prefer a car company that actually understands how to make and sell millions of safe and reliable products. Google has has a history of failing when it comes to shipping high quality hardware products.
I don't get the point of this article. I like the idea of tech companies expanding to other businesses. They can bring a fresh look to onboard systems, and make them less complicated and more elegant.
Yet the only thing they shouldn't forget is that cars are not computers. I'm not sure that Google's modified Prius will be as enjoyable (when driving) as a BMW 1 Series for instance.
The point of the article was a counter weight to Chunka Mui's on the same day which claimed Google was going to dominate a trillion dollar market - they will be a supplier at best with competitors, over a very long roll out period.
The title should be, Don't forget Volvo, Nissan, BMW and others are working on driverless cars too. The future will be awesome.
Why is anyone getting worked up over this.