Cruise founder Kyle Vogt takes OpenAI CEO Sam Altman on a virtual 75-minute drive in a driverless car through the streets of San Francisco in the video below. Cruise vehicles use LIDAR, radar, and optical sensors and are connected to to Cruise offices through the 4G cell network. The video demonstrates that autonomous vehicles are ready to enter ride hailing service, but not quite ready to go on sale to consumers.
To save time, the video has been speeded up five times so that 75 minutes are compressed into 15. During this time, the car handily deals with pedestrians, cyclists riding the wrong way, unprotected left turns, double-parked cars, and other road hazards. Vogt notes that the car’s computer not only tracks every other vehicle and pedestrian in its view, it simulates that vehicle or person’s movement and tries to predict where they are going so as to avoid any collisions. Since the accompanying map shows that the vehicle is tracking dozens of other moving objects at any given moment, this is pretty impressive.
Yet there were some problems that show why Cruise and parent GM won’t be selling driverless cars this year. At 3:30, the vehicle encounters a construction crew blocking an entire lane of a two-lane street and forcing it into the on-coming lane. The vehicle pauses as it tries to decide what to do. Vogt explains that the “AV got a little confused there” and consulted the “remote assistant system.” “It took like two clicks for that person to plot a path” and then the vehicle was on its way again.
In other words, there may be no driver behind the wheel, but someone in Cruise’s offices is monitoring everything the car does. When the vehicle encounters something it can’t handle, it signals for help. This happens, Vogt estimates, once every 5 to 10 miles, allowing one remote assistant to monitor several vehicles at a time.
best price on levitra If you are also dealing with sexual issues, then try not to solve the problems by yourself. generic cialis overnight Even the first mild attack of pancreatitis is an alarm bell of possible spasm of the sphincter. The duration of copulation increases after taking the tadalafil in canada drug. It is certainly worried that almost 50% male over 45 are suffering from impotence epidemic. generic viagra order
While this may work for robotaxis roaming carefully mapped cities, it isn’t a viable model for auto sales. It is one thing for GM and other auto companies to pay for an emergency-service staff for new-car buyers (which requires someone to pay an annual subscription fee of $150 a year or more), but a monitor who has to intervene every five to ten miles would be too expensive for most auto-buyers to pay for.
The accelerated video speed obscures the fact that the vehicle sometimes responded to unusual situations, such as a street that had perpendicular parking, by driving so slowly that other cars passed it, no doubt with some annoyance. The autonomous vehicle also seemed to leave more space between it and the vehicle in front of it than nearby human-driven vehicles, which could contribute to congestion by allowing fewer vehicles on the street at any given time.
For the most part, however, Cruise’s vehicle integrated well into San Francisco traffic. Vogt and Altman noted as well that the quality of autonomous vehicle software was growing “exponentially,” so in a few years computers will be much better drivers than humans. At that point, they suggested, the industry will probably focus more on reducing costs than improving the quality of the software.
Until computers master when it’s okay to not come to a complete stop at a stop sign, their driving won’t be better experience.
And that won’t be an issue until they figure out how to do Lidar on the super cheap. That won’t be easy. Hell, it may never happen.
Right now all this talk about eliminating a driver means they’re looking to spend $150K on a car just to save $50K on labor.
prk166,
You can buy an iPhone for about a thousand dollars that has LIDAR built in. Adding the LIDAR didn’t increase the cost of the phones over the previous editions. It sounds to me like the problem of super cheap LIDAR has already been solved.
Tough driving environment and it reminds me why I don’t relish driving in San Francisco. Certainly driving in a less dense suburban environment will be easier for an AV.
It seems to me that the best initial ROI for AVs is long-haul interstate driving. In the last few minutes they talk about platooning and how speeds can be increased and new road construction reduced if there was effective car-to-car communication and no human drivers. That’s looking pretty far into the future for dense cities but car or truck platooning on interstates could happen now. I would be happy to join an AV platoon in Denver and allow my car to drive me to Chicago. And if I wanted to pull over for a break, I imagine my car could join up with a different platoon that comes by. AV truck freight platooning wouldn’t need breaks except to refuel. That would lead to pretty substantial cost savings for freight companies.
Thank you, Mr. O’Toole.
I could swear that 2 years go they were talking $100K for lidar.
Do we know they’ve solved that problem? Having a cell phone that uses a lil laser to beef up photos for a subject a coule feet away is different that 360 scan for objects hundreds of feet away.
If they’ve solved it, great. I wouldn’t assume it’s been solved just cuz a cell phone “has it”. Those quotes are cuz for things like GPS, a cell phone “has it” but you wouldn’t want to do anything important using it.
Lidar used to cost $75,000—here’s how Apple brought it to the iPhone
I found Apple’s announcement particularly interesting because I’ve been working on a story about companies that are using the same combination of technologies—VCSEL lasers and SPAD detectors—to build much more powerful lidar for the automotive market. One of the big selling points of VCSELs and SPADs is that they can be created using conventional semiconductor fabrication techniques. As a result, they benefit from the huge economies of scale in the semiconductor industry. As VCSEL-based sensors become more common, they are likely to steadily get cheaper and better.
Two of the companies working on high-end VCSEL-based lidar—Ouster and Ibeo—have already gotten more traction than most companies in the crowded lidar business. Apple’s decision to adopt the technology—and the possibility that other smartphone vendors could follow Apple’s lead—will provide them with a nice tailwind in the coming years.
…
Ibeo, Sense, and Ouster are all rolling out new, low-cost designs because they expect an explosion of demand from the automotive industry. Lidar sensors could dramatically improve vehicles’ advanced driver assistance systems (ADAS).
For example, many people see Tesla as having one of the industry’s most advanced ADAS systems. But the company has a persistent problem with its vehicles crashing into stationary objects—occasionally with fatal results. Lidar is better than cameras or radar at detecting stationary objects, so adding lidar to cars could prevent many of these crashes while making ADAS systems more convenient for drivers.
Until now, lidar was considered too expensive for the automotive market, but that has started to change, with multiple companies promising lidar sensors that cost less than $1,000 in the next few years.
Ouster is aiming to have its ES2 sensor ready for mass automotive production in 2024. The company says it will initially cost $600 in volume, with the price falling to $100 in subsequent years.
I don’t know the specifics about the technology inside of the iphone vs. the technology inside of a self-moving car, but something tells me the requirements for LIDAR to help you take a selfie from 3 ft. away are much different than the requirements for LIDAR to constantly monitor all objects within x hundred feet of a moving vehicle.
True for any technology I suppose. The laser rangefinder that controls the cruise control on my 15-year-old automobile no doubt has very different requirements than laser rangefinders that provide targeting for laser-guided bombs and missiles.