If there was any doubt self-driving cars were coming and you were at NVIDIA’s Developers’ Conference this week, that doubt was put to rest.   Toward the close of the presentation, NVIDIA’s CEO gave a command and the Audi A7, reminding me a bit of KIT the car in the old Night Rider series, lit up and silently drove itself on stage.   This was ostensibly to showcase the power of the new NVIDIA Tesla K1 processor, the first mobile processor to have full desktop graphics technology and it showcased how far self-driving cars have come.

Self-Driving Revolution

Here in the Silicon Valley, there isn’t a day that goes by when we don’t see the goofy Google self-driving cars on the road.   These are mostly Lexus SUVs (some are ugly Toyota Prius cars) with a rotating LADAR system on the roof either in test or driving a Google employee to work.   The issue with these cars is that much of the inside is taken up with the self-driving technology and, on the outside, the cars don’t look cool - they look goofy and stupid - with all respect to Goofy.   People just aren’t going to buy a big car that drives but has very limited capacity and looks stupid to friends and neighbors. 

Well the most obvious revolution to the self-driving concept is what Audi and NVIDIA showcased - a self-driving core component that didn’t fill the car or even the trunk. At about 2 paperback books in size, it hides out of the way in the trunk and you don’t even know it is there.   In addition, they have found a way to embed the sensors behind car panels so, on the outside, the car looks no different than any other car. 

They picked the A7 to do the test.  This Audi, which ironically NVIDIA had loaned me several years ago, is drop dead gorgeous and I think most folks would be proud to own it.   Contemporary, fast and clean with not obvious way to tell that it was self-driving - except that it likely won’t ever get a ticket. Tthis was a far cry from the screwy looking Lexus SUVs that we are learning to dodge here in the valley.  

Image via Shutterstock.

Timing

Interestingly, there was also a TED event last week where one of the ex-Disney engineers focused on what was needed for self-driving cars.  The reason he is relevant is that Walt Disney had a vision decades ago with regard to this technology, in fact he had planned to turn EPCOT Center into a city but, unfortunately, that vision passed when he did.   This engineer’s list included knowledge of location (which you can do with GPS today) and knowledge of road rules (which can be provided by an online service). 

They need to be connected to each other to anticipate problems other cars are having and so the entire traffic system actually works as a system.   He thinks they will first be allowed to drive in HOV lanes to contain the risk and cars need to recognize between objects to make them safer - which is where the NVIDIA Tegra K1 comes in.  He also apparently talked about the need to redesign cities to make self-driving cars better optimize their efficiency which, if that happens, might revisit Disney’s vision for the city of tomorrow. 

Wrapping Up:  Oh My…

The speed we are moving to create self-driving capability is astounding.  We have jumped generations of technology both in terms of cars being able to see and recognize things and in making the core technology small enough to fit behind a panel in the back of a car and not take excessive room or power.   The day when we’ll be able to grab a good book, well eBook, and tell the car where to go with the confidence that your car will get you to that location quickly and safely is rapidly drawing near.  If you’d been at the conference you would have seen how near and likely, as I now do, have wanted an Audi A7 as well.   I am so looking forward to the end of traffic.