There are multiple approaches to tackling self-driving; one is to program algorithms or rules that will tell a car how to behave in specific situations. Nvidia is using a deep learning approach, however, by providing its autonomous system with real-world data from humans drivers and letting it learn how to drive on its own – like a supercharged, AI-powered teenager getting behind the wheel using only their experience of being a passenger to guide them.
A Google self-driving Lexus RX 450h was involved in a crash with a van in Mountain View, Calif. on Friday afternoon, according to local police. Thankfully, nobody was injured in the accident.
TechCrunch has reached out to Google for more information.
This week, Uber started picking up passengers with self-driving test cars in Pittsburgh – so it’s not surprising that just a few days later, Lyft has something to say on the subject. Lyft’s response may not be autonomous cars in active duty, but it is a lengthy, detailed treatise on how Lyft sees the future of transportation over the next ten years, as penned by Lyft co-founder and President John Zimmer.
“Parker, take the wheel.”
NVIDIA debuted its Drive PX2 in-car supercomputer at CES in January, and now the company is showing off the Parker system on a chip powering it. The 256-core processor boasts up to 1.5 teraflops of juice for “deep learning-based self-driving AI cockpit systems,” according to a post on NVIDIA’s blog. That’s in addition to 24 trillion deep learning operations per second it can churn out, too. For a perhaps more familiar touchpoint, NVIDIA says that Parker can also decode and encode 4K video streams running at 60FPS — no easy feat on its own.
Part of the arrangement that will bring self-driving Uber on-demand vehicles to Pittsburgh roads by the end of this year will also help Volvo bring its own self-driving cars, for use as either personal vehicles or as autonomous taxis, to market by 2021, according to the WSJ. The two companies are investing a combined $300 million, roughly split between the two, in a deal that will seek to achieve production of a road-ready autonomous car based on the XC90 SUV platform.
While the US waits to get Super Cruise in Cadillacs next year, Toyota has already rolled out a pretty robust V2V, or vehicle-to-vehicle, system in three models available in Japan. The latest version of the Prius, the Lexus RX, and the Toyota Crown, a luxury sedan sold in Japan, all have the ITS Connect system available as an option.
The cars communicate using a channel abandoned by analog TV transmissions when they switched to digital a few years ago. The cars send and receive very simple–but very useful–information: location, speed, and heading. “People ask about privacy” regarding these communications said Toyota spokesperson Hideki Hada in a phone interview, “but this is very obvious information.”
You could see 100D versions of the Model S and Model X in the near future.
Tesla has dropped hints that it’s ready to extend the range of its electric cars (there was an allusion to a “P100D” hidden in firmware), but it now looks like that long-distance technology is getting closer to fruition. Dutch regulators have approved 100D and P100D versions of both the Model S and Model X, hinting that a 100kWh power pack might soon hit the streets. If the listings are accurate, the Model S would get a whopping 380 miles on a charge — no mean feat when the 90D can ‘only’ manage 294 miles. The Model X would likely have a shorter range given that the existing SUV officially tops out at 257 miles, but it’s reasonable to say that you’d get over 300 miles on a charge.
There’s no official word on when you’d see 100D variants on the street, let alone how much they’ll cost. They’ll likely make the 60D seem like a bargain. Whatever price they’ll carry, they’ll bring Tesla one step closer to its dreams of cars that can drive cross-country (eventually, by themselves). And if history is any indication, the higher battery capacity will come along with a performance boost — add Ludicrous Mode and both EVs could put even the better supercars to shame in terms of short-distance acceleration.
It reportedly decided that it’s cheaper to buy EV power from someone else.
Nissan has long made its own electric car batteries thanks to Automotive Energy Supply, its team-up with NEC, but it appears to be having second thoughts. Sources speaking to both Nikkei and Reuters understand that Nissan plans to sell its controlling stake in AES, with NEC likely following suit. It’d just be less expensive to buy batteries from an outside supplier, according to tipsters. While it’s not certain who would snap up the business, the car maker is supposedly in talks with both Panasonic and “overseas companies” that include Chinese firms.
Samsung wants to make cars. Or at least, make parts of cars, in a similar supply-side capacity to the position it currently occupies in the mobile device and television world. A new Bloomberg report suggests Samsung is in talks with Fiat Chrysler to acquire some or all of parts-making subsidiary Magneti Marelli, which provides lighting, in-car entertainment and telematics to OEMs for use in their vehicles.
Comma.ai, the startup that George Hotz (aka Geohotz) founded to show that making driverless vehicles could done relatively cheaply using off-the-shelf components and existing vehicles, has open-sourced a dataset of 7.25 hours of highway driving.
It might not seem like a lot, but in terms of comparative datasets for highway driving out there, it is. And it’s what Hotz used to build the initial successful self-driving demo used to ferry Bloomberg around for comma.ai’s big public debut.