Waymo announces 7 million miles of testing, putting it far ahead of rivals
https://ift.tt/2M3ItEr
A Waymo executive announced on Tuesday that the company’s self-driving car fleet would reach 7 million miles of testing this week.
“We’ve amassed close to 7 million miles—we’ll hit 7 million miles this week,” said Shaun Stewart, Waymo’s chief business development officer, at the Innovfest Unbound conference in Singapore.
What makes this truly remarkable is that Waymo announced its last milestone—6 million miles—less than a month ago. Waymo CTO Dmitri Dolgov made that announcement at the Google I/O conference on May 8.
Dolgov didn’t specify exactly when Waymo reached the 6 million-mile mark. So the last million miles might have actually taken a bit more than a month to rack up. Still, Waymo’s pace of testing is clearly accelerating as the company gears up to launch its driverless taxi service later this year.
Waymo made this chart back in February to mark 5 million miles of testing. It shows that it took Waymo (then Google) more than a year to get from 1 million to 2 million miles. Then it took about six months each to get to 3 million and 4 million miles. Waymo logged its next million miles in about three months.
Now, three and a half months later, the company has racked up 2 million additional miles. And while Waymo made a big deal out of its early milestones, the last two have been low-key affairs, with executives casually dropping the new statistics in the middle of larger talks about Waymo’s plans.
No one else in the industry is close to matching the scale of Waymo’s testing. Uber had logged roughly 3 million miles before it was forced to suspend testing in March due to a fatal crash in Tempe, Arizona. GM’s Cruise drove only 131,000 miles in California between December 2016 and November 2017—the most recent period for which we have data.
Of course, miles on their own don’t prove anything about vehicle safety, as Uber’s deadly crash demonstrates. In an October blog post, Cruise CEO Kyle Vogt argued that testing in dense, chaotic San Francisco allows Cruise to learn more from each mile of testing than rivals like Waymo that predominantly test in suburban areas.
But Waymo’s accelerated testing program is just one of several signs that the Alphabet company is investing heavily to maintain its early lead in driverless vehicle technology.
Last month, Waymo announced that it had ordered 62,000 Chrysler Pacifica minivans, on top of the 20,000 Jaguar I-PACE order Waymo announced in March. Waymo’s most recent order is by far the largest such order in the driverless car business. The closest comparison is to Uber’s order of 24,000 cars from Volvo. But with Uber’s testing now grounded, it’s not clear if Uber will be ready to use those vehicles when Volvo is ready to start delivering them next year.
Waymo says it’s planning to launch a commercial driverless taxi service in the Phoenix area before the end of the year.
Machines that suck CO? from the air might be cheaper than we thought
https://ift.tt/2JAFpB5
When you spill a drink, you don’t say, “Oh well, the only thing we can do is spill fewer drinks in the future.” You grab a towel. So there’s also a natural attraction to the idea that we should develop a towel that can remove CO2 from the atmosphere. That isn’t as simple as grabbing one from a Home Goods store, however, and cost estimates have not fueled optimism for most methods of doing this.
Reforestation is an obvious option, but its potential impact is probably smaller than you think. Other biological schemes could include growing biofuels to burn in power plants that capture emissions and store them underground. Recently, we’ve also seen a couple of working pilot projects that look like a power plant run in reverse—they suck in air and harvest concentrated CO2, ready for storage.
One of those plants, located an hour north of Vancouver, British Columbia, is the brainchild of a company called Carbon Engineering. One of the founders of Carbon Engineering is Harvard’s David Keith, a researcher studying this and other conceivable methods of “geoengineering” our planet’s climate. This week, the Carbon Engineering team has published a nuts-and-bolts breakdown of its design, providing the first cost analysis of a working carbon capture plant.
Carbon Engineering’s pilot plant and another pilot plant built by Climeworks in Switzerland diverge at the substance used to grab CO2 from the air. Climeworks opted for solid granules of amine, while Carbon Engineering is using a water-based solution of potassium hydroxide.
Cycling components
In their paper, the researchers explain their design choices. While their end task is obviously different, they were able to adapt existing machinery for each step in the process. By getting a little creative and optimizing components, they were able to bring the cost down significantly from previous estimates. All of this is laid out in great detail.
The process starts with what is basically a cooling tower, where a fan pulls air in through a sort of filter that has a continuous flow of water trickling downward through it. But instead of cooling the water, the goal here is to get air in contact with the potassium dissolved in the water, which forsakes its hydroxide partner to grab molecules of CO2. The team designed a different geometry for the parts the water trickles through in order to maximize the chemical reaction, but otherwise this machine is essentially stock.
The potassium carbonate that forms enters something called a “pellet reactor,” more commonly used for special water treatment systems. The solution flows upward through tiny pellets of calcium carbonate, keeping them suspended. Add in a little calcium hydroxide, and the disloyal potassium trades partners again. The calcium carbonate that results causes the pellets to grow until they’re large enough to sink downward despite the flow. Potassium hydroxide flows out the other side, ready to get sent back to the first stage and grab another CO2 dance partner.
The enlarged calcium carbonate pellets now get sent to the third and final stage: heating that separates them into pure CO2 gas and calcium hydroxide that can go back to stage two. The heat comes from a natural gas furnace that produces steam. The steam travels through loops that efficiently dump heat into the calcium carbonate before reaching a generator turbine that provides the plant’s electricity.
I know what you’re thinking—they’re burning fossil fuel to capture CO2? While it’s possible to redesign this to run all-electric, in this configuration the CO2 emissions from burning natural gas are also captured, so this is not self-defeating from a climate standpoint.
Carbon neutral fuels
At this point, the pure CO2 could be compressed into a pipeline and sent off to a facility that pumps it underground for storage, reducing the amount in the atmosphere. But Carbon Engineering also has some other ideas. If you combine this with hydrogen gas split from water, you can synthesize conventional fuels.
Fossil fuels are made of carbon pulled from the atmosphere by plants millions of years ago and slowly cooked deep underground—carbon that we return to the atmosphere when fuels are burned. If you take your carbon from the atmosphere today and put it back tomorrow, you have no effect on the concentration of greenhouse gas. It’s the industrial equivalent of growing biodiesel. For transportation that resists electrification—notably air travel—that could provide a way to eliminate the greenhouse gas emissions.
Unless you’re aiming to build your own hobbyist carbon-capture plant, the biggest thing here is probably the cost estimate. While this pilot plant can currently only capture about 200 tons of CO2 per year, Carbon Engineering’s estimate is based on a larger “commercial-scale” plant design that would handle about 1 million tons per year—equal to the emissions of roughly 200,000 US cars. Scaling up is simple for the modular components (just stack ‘em), but a couple of pieces would become more cost-effective when supersized.
The end result is a cost of $94 to $232 per ton of CO2 captured by a plant of that size. Past estimates for capture plants have been all over the place but are generally much higher than that. A major 2011 report, for example, estimated a plant like this would cost at least $550 per ton.
For comparison, British Columbia’s tax on emissions is currently CAN$35 per ton of CO2, rising to CAN$50 in 2021. That’s generally the range of existing carbon prices, though there have certainly been arguments for higher numbers. If a carbon capture plant could come in below $100 per ton—and the Carbon Engineering folks certainly don’t think they’re done cutting costs—it could enter the conversation with other plausibly economically viable technologies for countering emissions.
Of course, not spilling a drink is still cheaper than buying a towel to clean it up.