It is by now pretty much a given that the mania to cut taxes to enrich the wealthy has resulted in government being deprived of the necessary revenue to maintain its existing infrastructure and public spaces, let alone make any improvements to bring them into line with other developed nations. Anyone who has traveled to other developed and even many developing countries will immediately notice the difference in roads, airports, and other transportation systems and how the once-enviable public spaces of the US, such as its roads, parks, and libraries, are slowly decaying.
It turns out that the poor condition of roads in the US is hurting the development of self-driving cars.
Volvo’s North American CEO, Lex Kerssemakers, lost his cool as the automaker’s semi-autonomous prototype sporadically refused to drive itself during a press event at the Los Angeles Auto Show.
“It can’t find the lane markings!” Kerssemakers griped to Mayor Eric Garcetti, who was at the wheel. “You need to paint the bloody roads here!”
Shoddy infrastructure has become a roadblock to the development of self-driving cars, vexing engineers and adding time and cost. Poor markings and uneven signage on the 3 million miles of paved roads in the United States are forcing automakers to develop more sophisticated sensors and maps to compensate, industry executives say.
Tesla CEO Elon Musk recently called the mundane issue of faded lane markings “crazy,” complaining they confused his semi-autonomous cars.
The article goes on to explain what information the cars need to be able to navigate safely on public streets. About 65% of US roads are in poor condition, an astonishingly high number. In order to compensate for that and the lack of uniformity across the nation in how road markings and traffic lights are configured, cars that will be driven on US roads will have to incorporate much higher levels of technology to compensate, making them more expensive.
Obviously taxes haven’t been cut quite enough yet to trigger that magical increase in government revenue.
And that was in Los Angeles! Think how much trouble he would have in a more northern latitude, where the road markings are regularly “sand-blasted” by snow plows and road salt.
I get visions of Henry Ford’s Model T, which was designed to withstand unpaved dirt and mud roads of the 1910s. US urban road conditions are returning to the level of rural roads from 40 years ago.
A related news story is Washington DC’s metro rail system. The tracks are in such states of disrepair that major lines will have to be closed for six months.
Is this what it will take for the US’s car cult (it’s not “car culture”) to come to an end? Not the end of cheap gasoline, but the end of drivable roads and bridges? Maybe US politicians will finally encourage people to use bicycles. At least mountain bikes could traverse the roads and the lighter loads won’t cause bridges to collapse. It is summertime, after all, so the weather’s agreeable.
Try riding a bike in Texas during the summer.
If the self-driving cars have trouble seeing navigation markings, so do less-experienced drivers (read: teenagers) and anyone with impaired vision. Crumbling infrastructure makes us all less safe. And that’s ignoring the occasional catastrophic bridge failure. One of my least preferred ways of dying is via drowning, but a major river cuts through my city and I cross it often on one of the 10 major bridges that connects my town. Every time a bridge collapses in another state, I get a little more nervous. Our worst bridge is being replaced even as I type, but that’s only one.
Looks like it’s time to roll out Flying Cars!!
@Jockaira #6, sounds great. Let me know when they can replace driving over bridges, I’ll stand in line for a week to buy a car like that.
While the level of disrepair that the US government has allowed is atrocious, automated cars must be able to handle things like faded lines and bad signage if they are ever to be viable. They must be able to handle the worst case (or close to it) or they are worthless. Even if 95% of the roads were in perfect condition, what do you do if your destination is in the remaining 5%? What if the road gets covered with snow/puddles/sand/mud? Google already wants to release cars with no driver controls but they have been hindered by regulations (I agree with the regulations at this time).
One of the solutions to this is the inclusion of LIDAR and robust 3-D maps (the Google cars have this). LIDAR is still pretty expensive but the price is dropping steadily. Musk has stated that he didn’t think LIDAR was required for fully-autonomous cars and this problem is precisely why I disagree. Good luck driving in any northern climate if the vehicle needs to see the painted lines on the road.
Re flying cars: While I think they would be awesome, I really hope this doesn’t happen any time in the near future, especially if humans are in control. You wouldn’t be safe anywhere; you could get hit by a car while sleeping in your 2nd floor bedroom.
Golgafrinchan, even if automation doesn’t handle faded lines or bad signage at the level it handles good infrastructure, that doesn’t necessarily make it worthless. Simply compare their rate of accidents (or other more complex measures of harm) in such conditions to actual drivers. My understanding is that googles algorithms would make driving safer in a utilitarian sense, in California, at least.
A sleep-deprived me certainly has trouble handling such conditions, as someone else pointed out regarding unexperienced drivers.
Algorithms don’t need to be perfect to better than humans.
@ Sam N #9,
Yes I totally agree that they will be safer. They already appear to be better than human drivers and they’ll continue to get better. I just meant they’ll only be good for small niche applications if there are a significant number of roads that they don’t work on, and that will always be the case if they rely primarily on cameras detecting lane markings. Yes, “worthless” was an overstatement but they won’t be able do do away with user controls* outside of locations with the most ideal climates (as Google attempted to do before Califonia’s recent law on the subject).
Adding LIDAR to the sensor suite (GPS, WiFi location, radar, cameras, accelerometers, digital gyros, wheel odometers, V2V systems) removes that obstacle because the 3-D maps include the positions of all physical objects (railings, curbs, bridges, boulders, trees, signposts, ditches, etc.). Also, LIDAR is second only to radar in robustness in extreme weather.
However, LIDAR is still a bit pricy for widespread use. The 64-beam LIDAR used on Google cars costs around $75,000 but, since they are being used to create the 3-D maps, they need to be a bit more accurate. There are simpler units in the low thousands of $, but that’s still too expensive for mass adoption. The prices are dropping rapidly so they should be more reasonable soon.
*Keeping user controls and requiring deivers to take over occasionally is problematic because people have proven to be unreliable at quickly taking over. Even the Google test drivers haven’t been great at quickly taking over and they’re trained to do it.