Waymo just dropped a 43-page white paper called the Waymo Safety Report that provides a wealth of new details about Waymo’s vision for the self-driving car product the company is getting ready to launch.
Officially, the document is a regulatory filing with the National Highway Traffic Safety Administration, which has encouraged— but not yet required —the makers of self-driving cars to file a report describing how they expect to deal with a variety of safety issues. But the document is also another part of the public education campaign the company has been running to convince Americans of the benefits of its technology.
It’s fundamentally a marketing document rather than a technical one, so it leaves a lot of unanswered questions about exactly how Waymo’s technology will work. Still, it provides a lot of new information—and publicly confirms a lot of rumors and educated guesswork—about how Waymo envisions the self-driving car product that Waymo could launch as soon as this year.
Here are five big things I learned from reading the report.
1) Waymo is putting safety at the center of its pitch
The biggest fear with the advent of fully self-driving cars is safety: what happens if a car malfunctions, crashes, and kills passengers or bystanders?
Waymo’s response is to put safety at the heart of its pitch for driverless vehicles. “Safety is at the core of Waymo’s mission,” the report says. “It’s why we were founded over eight years ago as the Google self-driving car project.”
The report notes that 1.2 million people around the world are killed by car crashes every year and says that 94 percent of those crashes are due to human error. While there are inherent risks to any new technology, Waymo argues that at scale, its technology will be able to save thousands of lives every year.
2) Waymo says its cars can handle nighttime driving and light rain, but its cars will be geofenced
We’ve known for a while that Waymo was planning to offer a ride-sharing service rather than selling Waymo cars outright to customers. This latest report confirms that plan and fills in some details on how it will work.
One big advantage of the ride-sharing model is that it isn’t necessary to master every possible driving scenario before beginning to offer service to the public. Waymo plans to limit customers to trips in areas where Waymo has detailed maps.
“Passengers cannot select a destination outside of our approved geography, and our software will not create a route that travels outside of a geo-fenced area, which has been mapped in detail,” Waymo says. The Information reported earlier this month that Waymo’s cars will initially be made available in the area of Chandler, a Phoenix suburb.
Similarly, if the technology isn’t ready to handle a certain weather situation—say, snow and ice—a company can initially just not offer service in those conditions. This is likely a big reason Waymo is choosing to launch in Phoenix first. In Phoenix, people don’t have to worry about snow or ice, and even rainstorms are rare. Still, Waymo says that its cars are designed to operate in a fairly wide range of conditions.
“We’re developing self-driving technology that can navigate city streets in a variety of conditions within broad geographic areas,” Waymo writes. “Our vehicles are designed with the capability to drive in inclement weather, such as light to moderate rain, and can operate in daytime and at night.”
If the weather suddenly takes a turn for the worst, Waymo cars are programmed to detect this, find a safe place to stop, and pull over until the weather improves.
3) Waymo is sandboxing safety-critical systems and building in a lot of redundancy
If your smartphone crashes, you reboot it and carry on with your day. If the computer on a self-driving car crashes while the car is on the freeway, the consequences could be deadly.
Waymo says it has done extensive work to make sure that computer crashes don’t lead to car crashes. All of the key systems on its cars—the computer, brakes, steering systems, and batteries—have backups ready to take over if the main system fails.
“A secondary computer in the vehicle is always running in the background and is designed to bring the vehicle to a safe stop if it detects a failure of the primary system,” Waymo says. Waymo also says that there are “independent collision avoidance systems” on board the cars that scan the road ahead for obstacles and can slam on the brakes if they see a pedestrian or other obstacle the main computer missed.
According to Waymo, all of these systems have separate power sources, so the failure of power to the main computer won’t knock out the backup computer or other critical systems.
Waymo says it has thought hard about computer security, too.
“Safety-critical aspects of Waymo’s vehicles—e.g. steering, braking, controllers—are isolated from outside communication,” Waymo writes. “For example, both the safety-critical computing that determines vehicle movements and the onboard 3D maps are shielded from, and inaccessible from, the vehicle’s wireless connections and systems.”
The obvious question here is how these safety-critical systems manage to get necessary information from the outside world—like updates about traffic conditions or even just information about where the customer wants to go. I’ve asked Waymo for more details and will update if they tell me more.
4) Waymo cars will feature a "pull over" button and one to call customer service
Waymo says it will have a ride-sharing app that customers can use to hail a car just as they do with Uber or Lyft today.
There will also be controls inside the vehicle. A display screen will show “important trip information such as destination and time to arrival. It also visualized static road elements like traffic lights, stop signs, and dynamic agents in the environment such as vehicles, cyclists, and pedestrians.”
The goal here is not only to inform customers of the ride’s progress but also to persuade customers about the safety of the technology. Nervous passengers will be able to verify that the car sees the same pedestrians, stop signs, and other key objects that the passenger sees out the window.
Since the car will drive itself, there’s minimal need for controls inside the car. But there will be a couple of key buttons. One will be a “pull over” button the passenger can push any time they want to get out of the car. When pushed, this button will “identify the nearest location to safely pull over so that the rider can exit the vehicle.”
Another button will patch passengers through to Waymo’s operations center where they can talk to a human Waymo rep. Staffers at the operations center can answer customer questions and intervene if a car encounters problems. Customers will also be able to talk to a Waymo rep via the mobile app.
5) Waymo seems to be gearing up for commercial release
The larger lesson from this report is that Waymo has thought through a lot of details about how to turn a promising technology into a full commercial product.
Waymo says it has subjected its vehicles to a wide variety of extreme situations to test their durability. “We blast our components with ultraviolet radiation, bombard them with powerful water jets, dunk them in nearly freezing vats of water, corrode them in chambers full of salty mist, shake and shock them with powerful vibrations, and heat and freeze them for weeks at a time in temperature and humidity chambers,” Waymo writes. “We analyze any failures and make design improvements to increase the reliability of our components.”
Waymo has developed protocols for dealing with crashes and has begun training law enforcement officials in Chandler and elsewhere about how to interact with its vehicles.
Waymo launched a public education campaign to persuade customers that the technology will save lives and empower elderly and disabled customers.
These are the kinds of steps a company takes when a product is months, not years, away from commercial release. The Information has reported that Larry Page, the boss of Waymo CEO John Krafcik, has been pushing Waymo to launch a commercial product this fall. We don’t know if Waymo will hit that ambitious goal. But it’s clear that Waymo is aiming to bring its product to market soon—likely in 2018, if not this year.
- Apple Self-Driving Cars Latest News & Update: iPhone Maker Investing Big In Driverless Car Market? What We Know So Far
- Nvidia GTX Titan X Works Under Biggest Pascal Chip On GeForce Card; NVIDIA To Perfect Tesla Self-Driving Cars?
- Google Spins Off Self-Driving Car Unit, Now Called 'Waymo'
- Google Self-Driving Car Sticky Adhesive Will Save Pedestrian in Collision?
- Self-Driving Car ALVINN: 1989 Invention Acts As Precursor To Advanced Technology Behind Driverless Cars
- Nvidia Self Driving Car Release Date, News & Updates: Driver-Less Cars on California's Highways?
- VR and self-driving cars are a fascinating, scary combo
- Obama Outlines Self-Driving Car Guidelines
- Google-Chrysler Self-Driving Car Deal Almost Done?
- CES 2015 Preview: What To Expect From This Year's Show, Including Self-Driving Cars And Bigger, Better Displays
- Uber Self-Driving Vehicle Banned On California For Not Acquiring Testing Permit; Google's Self-Driving Cars Set To Outpace Uber?
- Nvidia Announces New AI Supercomputer for Self-Driving Cars
- Google's Self-Driving Cars Will Speed If Necessary
- Apple Finally Confirms Self-driving Car Speculations
- Google Self Driving Car Revealed, Does Not Include Steering Wheel Or Human Controls [VIDEO]
- ‘Google Car’ Release Date, News & Update: Google Abandons In-House Self-Driving Car Project; Partners Fiat Chrysler, Waymo For 2017 Launch
- Apple Latest News & Update: iPhone Maker Plans To Build Self-Driving Car; Donald Trump To Review Regulations?
- Apple Forms Secret Team to Experiment With Electric Cars - Report
- This hard drive reliability report highlights why it’s important to back up your data
- What We Learned From EA’s Q3 Financial Report
5 things we learned from Waymo’s big self-driving car report have 1654 words, post on arstechnica.com at October 13, 2017. This is cached page on Talk Vietnam. If you want remove this page, please contact us.