Day: February 3, 2020

Inclement conditions are challenging for autonomous vehicles for several reasons. Snow and rain can obscure and confuse sensors, hide markings on the road, and make a car perform differently. Beyond this, bad weather represents a difficult test for artificial intelligence algorithms. Programs trained to pick out cars and pedestrians in bright sunshine will struggle to make sense of vehicles topped with piles of snow and people bundled up under layers of clothing.

“Your AI will be erratic,” Czarnecki says of the typical self-driving car faced with snow. “It’s going to see things that aren’t there and also miss things.”

Matthew Johnson-Roberson, a professor at the University of Michigan who is developing a delivery robot optimized for difficult weather conditions, believes tackling bad weather may offer a way to gain a competitive edge. Troubling conditions are a major source of accidents, he says, so they arguably should be a priority.

“The really big players are not focused on this,” says Johnson-Roberson. “There’s still a lot of work to be done on self-driving cars in general, but [driving in bad weather] is going to be a big differentiator, and also important to scaling this.”

Waymo declined to comment, but a spokesperson pointed out that the company’s newest sensors and software are better suited to challenging weather. A spokesperson for Argo said initial deployments of the company’s technology should be able to handle light rain, but “for heavier rains and snow, there still needs to be advancements in both hardware and software.”

When industry players decide to tackle bad weather, they’ll gather lots of training data of their own. But in the meantime, Czarnecki’s data should help the field advance.

“Adverse weather presents tremendous challenges for automated driving technology, and I applaud these researchers for releasing a challenging-weather data set,” says John Leonard, a professor at MIT who is also affiliated with the Toyota Research Institute. “Publicly available data sets can have a huge positive impact on research in the field.”

“The complexity of winter weather is going to take an incredible amount of work for automation technology to tackle,” says Bryan Reimer, a research scientist at MIT specializing in autonomous driving. “Ice-weather conditions are incredibly difficult.”

As for tackling the worst conditions on, say, Interstate 70 through the Rocky Mountains, where 24-hour snow crews are required, Reimer thinks self-driving cars won’t be there for a while. “The only way you’re going to drive that autonomously is to heat the road,” he says.

More Great WIRED Stories

It’s perhaps the best known and more worrisome of climate feedback loops: As the planet warms, permafrost—landscapes of frozen soil and rock—begins to thaw. And when it does, microbes consume organic matter, releasing CO₂ and methane into the atmosphere, leading to more warming, more thawing, and even more carbon emissions.

But here’s something you’ve probably never heard of, and it’s something not even the UN’s Intergovernmental Panel on Climate Change has really considered: thermokarst. That’s the land that gets ravaged whenever permafrost thaws rapidly. As the ice that holds the soil together disappears, hillsides collapse and massive sinkholes open up. Climate scientists have been working gradual permafrost thaw into their models—changes that run centimeters deep over decades or centuries. But abrupt permafrost thaw happens on the scale of meters over months or years. That shocks the surrounding landscape into releasing potentially even more carbon than would have if it thawed at a more leisurely pace.

Today in the journal Nature Geoscience, researchers argue that without taking abrupt thaws into account, we’re underestimating the impact of permafrost thaw by 50 percent. “The amount of carbon coming off that very narrow amount of abrupt thaw in the landscape, that small area, is still large enough to double the climate consequences and the permafrost carbon feedback,” says study lead author Merritt Turetsky, of the University of Guelph and University of Colorado Boulder.

Less than 20 percent of northern permafrost land is susceptible to this kind of rapid thaw. Some permafrost is simply frozen rock, or even sand. But the kind we’re worried about here contains a whole lot of water. “Where permafrost tends to be lake sediment or organic soils, the type of Earth material that can hold a lot of water, these are like sponges on the landscape,” says Turetsky. “When you have thaw, we see really dynamic and rapid changes.”

That’s because frozen water takes up more space than liquid water. When permafrost thaws, it loses a good amount of its volume. Think of it like thawing ice cubes made of water and muck: If you defrost the tray, the greenery will sink to the bottom and settle. “That’s exactly what happens in these ecosystems when the permafrost has a lot of ice in it and it thaws,” says Turetsky. “Whatever was at the surface just slumps right down to the bottom. So you get these pits on the land, sometimes meters deep. They’re like sinkholes developing in the land.”

“Essentially, we’re taking terra firma and making it terra soupy,” Turetsky adds.

As the earth turns to soup, the landscape begins to scar. The process is so rapid and so violent, Turetsky says, that sometimes when she returns to a site she’s monitoring to check her temperature and methane sensors, she’ll find they are gone. “When you come back in, it’s a lake and there’s three meters of water at the surface. You have to probably say goodbye to your equipment,” she says.

When these lands thaw, they play host to a number of processes. As ice turns to liquid water, trees flood and die off. Thus more light reaches the soil, further accelerating thawing. This is in contrast to gradual thaw, when the plant community largely stays the same as the ice thaws. Defrosted soil at the surface gets thicker and thicker, but it doesn’t catastrophically collapse.