5 ms·
Google doesn't want to build cars. Google wants to be the Microsoft of cars, to relegate the Fords and GMs of the world to hardware-builders akin to IBM. It i
by sandworm101 11y ago
Google doesn't want to build cars. Google wants to be the Microsoft of cars, to relegate the Fords and GMs of the world to hardware-builders akin to IBM.
It is a mistake to try. As I've said in a hundred other threads, some basic autodrive functionality has been available for years and every focus group has dismissed it. Speed regulators, automatic enforcement of follow distance, the automatic seatbelt (which only existed to delay the airbag requirement) anti-distraction devices such as entertainment systems that lock down at speed ... all would be trivial to implement but have been summarily rejected by the general market. The concept that Google will somehow implement all of those things at once, and that people will buy in, seems fanciful.
I still haven't seen any safety numbers that account for the most dangerous driving hazard. I haven't seen any evidence that drunk people crash less often in an auto-drive car. All on the road so far have had driver overrides, which a drunk would of course be able to use. Nobody has dared to put drunks in charge of these things. So we do not know whether they will reduce drunk driving accidents or not. It might perhaps increase the issue if lots of drunks decide to rely on autodrive cars, creating a greater opportunity for them to wrest control from the robot.
And I've yet to see the google cars react to unusual situations such as a person directing traffic or boarding a ferry. Those driver overrides, which allow the drunk to take control, remain necessary.
- dclowd9901 11y agoI'm finding it very difficult to dig up actual statistics on what percentage of accidents each year are caused by drunk drivers, most likely because I suspect the percentage is extremely low. The nearest I could find was a statistic that about 7% of all vehicle accidents involved an alcohol impaired operator (http://www.encyclopedia.com/doc/1G2-3403100014.html http://www.encyclopedia.com/doc/1G2-3403100014.html). That said, why do you think impaired drivers are the biggest danger on the road? I firmly believe it is simply bad, irresponsible, untrained drivers caused by 1) the ease of obtaining a driver's license, 2) the lack of expense overall in owning and driving a car and 3) the very low standard of acceptable driving that's the most troublesome. It's going to be easy to throw a drunk in an autonomous vehicle and ship them home. It's going to be exponentially more difficult to provide access to autonomous vehicles to the general population.
- nabla9 11y agohttp://www-nrd.nhtsa.dot.gov/Pubs/812102.pdf http://www-nrd.nhtsa.dot.gov/Pubs/812102.pdf In 2013, 10,076 people were killed in alcohol-impaired- driving crashes, an average of one alcohol-impaired-driving fatality occurred every 52 minutes. These alcohol-impaired-driving fatalities accounted for 31 percent of the total motor vehicle traffic fatalities in the United States.
- dclowd9901 11y agoThis is a percentage of fatal accidents, which on the whole are very statistically unlikely (they account for less than 1% of all accidents). Just because someone didn't die doesn't mean an accident was inconsequential. In other words, if I told you 1/3 of fatal airline crashes were caused by drunken pilots but half of all airline crashes were caused by fatigued pilots, which would you try to fix first?
- sandworm101 11y agoFrom the CDC "In 2013, 10,076 people were killed in alcohol-impaired driving crashes, accounting for nearly one-third (31%) of all traffic-related deaths in the United States." "Drugs other than alcohol (e.g., marijuana and cocaine) are involved in about 18% of motor vehicle driver deaths. These other drugs are often used in combination with alcohol." http://www.cdc.gov/motorvehiclesafety/impaired_driving/impaired-drv_factsheet.html http://www.cdc.gov/motorvehiclesafety/impaired_driving/impai... So with 31+% of deaths related to alcohol, I'd put that as the greatest threat. The problem with all accident stats is that one accident creates many stats. Say a drunk hits a tree. One person says it was "speed related", another blames a distraction, another the lack of devices to prevent the car leaving the road, another the driver not using a seat belt or airbag ... and we all know it begins with the drunk. I eagerly wait for Google to allow drunks to use these cars. That "the car gets you home" scenario is what supporters champion when they talk of reducing accidents. But I doubt Google's lawyers will ever remove "do not drink and drive" from the car's manual.
- dclowd9901 11y ago
- kwhitefoot 11y ago> Speed regulators, automatic enforcement of follow distance, ... have been summarily rejected by the general market. Nonsense. Lots of new cars have cruise control with front facing radar distance and closing speed measurement that feeds back to the engine management to slow the car if it is getting too close too fast. Such radar is also used to apply the brakes in case the driver fails to brake soon enough. Here in Norway such facilities are available on small VW work vans as well as Toyota cars and of course luxury cars like the more expensive Mercedes, Jaguar, and Audi. Of course I agree with your final point that at the moment there are many circumstances in which a human driver must take control but there are many small measures that can be taken by the car makes to further reduce the frequency and severity of accidents by adding features like adaptive cruise control, lane sensing, pedestrian and other obstacle sensing and avoidance, etc. These will only be rejected when they are seen as extra cost add-ons but when they become common people will simply expect them to be fitted.
- sandworm101 11y ago>> Lots of new cars have cruise control with front facing radar distance and closing speed measurement that feeds back to the engine management to slow the car if it is getting too close too fast. Such radar is also used to apply the brakes in case the driver fails to brake soon enough. I didn't mean those. I mean mandatory devices, things that totally replace the driver's decision making. A speed regulator would limit the car to the speed limit no matter what the driver wants. Such a car would be electronically incapable of speeding. That isn't cruise control. Trucks have had such devices for years. And those radars for follow distance still allow drivers to tailgate. I'm talking about a robot stepping in to prevent bad behaviour with no option to override. That is what has been rejected by consumers year after year. But that is what google is pitching. Or how about: No seatbelt, no move. That is dead simple to implement, but no US manufacturer has even attempted to market such a device.
- Al-Khwarizmi 11y agoConsumer habits change. I'm sure some people who have driven all their life won't give up control voluntarily, but you know all those articles in The Atlantic, etc. about "millennials" that don't care for cars and driving anymore? This could be the automobile industry's chance to recapture their interest (and money). OTOH incremental changes, some of which (like speed regulators) only add restrictions but don't save work for the driver, are not comparable to the concept of the car driving itself and the rider sitting back and relaxing. If there were a true self-driving car in the market (and the legal hurdles were cleared) I don't see how it wouldn't sell like hotcakes.