Previous: Why Are Pandas Black and White?
Next: How the Electricity in Our Bodies Could Fight Cancer



View count:114,862
Last sync:2020-10-10 18:09
Self driving cars and self-repairing roads: the future of driving is bright, or at least less aggravating.

Hosted by: Hank Green

SciShow has a spinoff podcast! It's called SciShow Tangents. Check it out at
Support SciShow by becoming a patron on Patreon:
Huge thanks go to the following Patreon supporters for helping us keep SciShow free for everyone forever:

Bd_Tmprd, Jeffrey Mckishen, James Knight, Christoph Schwanke, Jacob, Matt Curls, Sam Buck, Christopher R Boucher, Eric Jensen, Lehel Kovacs, Adam Brainard, Greg, Sam Lutfi, Piya Shedden, Katie Marie Magnone, Scott Satovsky Jr, Charles Southerland, Charles george, Alex Hackman, Chris Peters, Kevin Bealer
Looking for SciShow elsewhere on the internet?
Links to Original Episodes and Sources:

Turning Your Leftovers Into Fuel

Are Electric Cars Really More Environmentally Friendly?
Why Are Self-Driving Cars Taking So Long?

Why Are Self-Driving Cars Taking So Long?

These Smart Roads Could Change the Future of Driving

 (00:00) to (02:00)

*intro music plays*
In 1908, Henry Ford introduced the Model T, and just like that driving became a major mode of transportation. We've come a long way since then, and the technology involved in driving from the fuel to the cars and even the roads we drive on continues to advance. We've talked about a lot of these advancements here on Scishow, so today we're gonna take a look at what the future of driving may look like. It took less than two decades to go from the first Model T to the 15 millionth one, and nowadays there are more than a billion cars driving around.  While that is amazing, in some ways its been pretty crummy for the planet because the vast majority of those cars run on fossil fuels. But many researchers are trying to change that so in the near future what you fill your tank with could be very different. Like you might be filling it with your table scraps. I'll let younger me fill you in on that. 
It's three in the afternoon on a Wednesday when you hit the afternoon slump, you're tired, and probably not just because you were up late binge-watching season 4 of Buffy the Vampire Slayer. Your body's low on energy so you grab an apple for a quick snack. Food is energy after all, thats why we eat it. But instead of throwing away that core when you're done, try putting it in the garbage disposal. There's a lot of energy left in there and this way it might be converted into something useful. Cause when food waste ends up in landfills it actually creates a lot of problems. It takes time for that uneaten apple core to decompose, and meanwhile, bacteria are turning it into methane and that's an issue because methane is a greenhouse gas and it's actually significantly better at trapping heat than carbon dioxide. Which got all the scientists thinking what if you could figure out a way to control how the food waste decomposes and in the process produce energy we can actually use. Add the potential to help the planet and makes lots of money so win win really, but it wasn't gonna be simple. The first challenge was just getting the food waste itself. Recycling programs are pretty successful but......

 (02:00) to (04:00)

it's still hard to get people to agree to scrape the food out of the takeout container before they throw it in the trash or get restaurants to separate food scraps. Cities are slowly starting to get on board mandating that commercial establishments like restaurants send no more thna  acertain percentage of their waste to landfills. But even after they've separated out the food waste, there's still the matter of getting it somewhere to be processed, not to mention figuring out how to process it. At first, cities decided to take advantage of existing sewage systems. We've here on SciShow about where the stuff in your toilet and sink goes after you flush and eventually, part of it ends up in an anaerobic digester. Basically just a giant tank full of sludge and some specially selected microorganisms, but very little oxygen. Bacteria would normally digest that sludge using a reaction that has carbon dioxide as one of the products. But making carbon dioxide requires oxygen and when there isn't enough around, they have to do the best they can with what they have. So they end up producing methane instead, and that's much more useful to us because the treatment plant can harness that methane and turn it into a useable fuel. When your apple core goes down the drain, it lands in the sewage system and eventually ends up in the sludge where bacteria can use it to produce even more methane. So at first, cities simply added their food waste to the mix. Once they figured out that this was an effective way to turn food waste into methane, companies started building treatment plants specifically designed to handle food waste. Many of these plants have deals with local utility companies that are able to sell the extra fuel in the form of heat or electricity or just compressed natural gas. Sell enough, and they turn a profit. But that still isn't an idea fuel. To move it around or burn it, you can to compress it first, which takes extra energy and makes the fuel less efficient. And while methane is easy to use in power plants once they get it, it's more difficult to use compressed natural gas to, say, power a car. That's why these days many companies are looking to turn food waste into an even more useful form of energy: ethanol. The stuff that is in your tequila! Ethyl alcohol. There are plenty of industrial plants that produce ethanol from things like corn, and they've been doing it for decades. One of the more common techniques uses enzymes to convert cellulose into sugar and then allows yeast to turn that sugar into ethanol and carbon dioxide. Since those plants use

 (04:00) to (06:00)

very specific raw materials or feedstock, the chemical reactions are easier to control. Producing ethanol from an unpredictable mixture like food waste is more complicated. But there are already bio refineries that use other kinds of waste to produce ethanol, like one in Edmonton, Alberta that started operations last year. That plant takes what's known as municipal solid waste, anything that can't be recycled or composted. Then it sorts that waste, separating out stuff that can be used as feedstock for that plant, like non-recyclable plastics and soiled cardboard. The feedstock is processed using gasification, where high temperatures, hot sand, steam, and oxygen are used to convert it into a mixture of carbon monoxide and hydrogen called syngas. Using catalysts to help spur the reactions along, the syngas is turned into chemicals like methanol and ethanol. Companies haven't yet applied the technology to food waste on an industrial scale but they're working on it.

 NewSection (4:51)

Using leftovers to power cars would certainly make them more efficient, but we could just use electric cars instead, since those are much more environmentally friendly than the gasoline kind. Or are they? You might have heard that electric cars are actually worse for the planets than gas-powered ones. But that is probably not true. Here's more on that. 

 NewSection (5:12)

Telling people that buying an electric car is a great way to fight climate change is a pretty reliable way to start an argument in some circles. People will say that you still burn fossil fuels with electric cars, it just happens at the power plant instead of the engine. And while that is at least partly true, over their lifespans, electric cars don't consume anywhere near the fossil fuels that gas-powered cars do, especially in the United States, with electricity getting cleaner all the time, they're even better than you might expect. There are a few variables to keep in mind, though. Starting with the cars themselves. In the US, as of 2016, the average pure-gasoline passenger car goes about 9 kilometers for each liter of gas it burns, or about 22 miles per gallon. But that's just an average. Some go 4 kilometers or so, others go 14. 

 NewSection (6:01)

 (06:00) to (08:00)

But the fuel efficiency hits higher highs and lower lows when we're talking about hybrid cars or trucks on the interstate. Plus, the act of manufacturing a car leads to greenhouse gas emissions whether it's electric or not. So does refining gasoline. The major complicating factor here is electricity, and where you live determines how clean your electricity really is. Most electricity in the US uses a combination of natural gas, coal, and nuclear fission, with a bit of water, wind, solar, oil, and a few others thrown in. But those numbers change from state to state and depend on things like local natural gas sources or how windy it is today. For example, Alaska has plenty of natural gas and hydroelectric resources, so its power plants very little waste when generating electricity. That means a full electric vehicle charge using Alaskan electricity creates roughly the same emissions as a gas engine that drives 48 kilometers per liter or 112 miles per gallon. That's roughly 5 times the national average and 2 or 3 times better than even some of the best hybrids. It's pretty efficient. The other extreme are places like Colorado, which energetically speaking, is one of the dirtiest states. Sorry, Colorado. About half of Colorado's electricity is from coal, which produces more emissions than just about any other source. But even there, electric cars still out perform gas cars. a full charge off of Colorado electricity equates to about 20 kilometers per liter, 46 miles per gallon. About double the national average based on 2016 figures. That's pretty good, better still, most of the country is closer to Alaska's numbers than Colorado's. it comes down to this: power plants are just better at making power than car engines are. One reason is that they're simply bigger. Bigger things don't waste as much energy staying hot and that makes them more efficient. The story is similar across most of the world, although again, the details change depending on

 NewSection (8:01)

 (08:00) to (10:00)

where your electricity comes from. In countries that tend to use more coal, like India or China, electric cars break even with the average gas-powered car in the US, although they are still less efficient than the average Indian gas-powered car. But in water-powered Paraguay, or geothermal-rich Iceland, gas engines need to get more than 90 kilometers per liter to beat an electric one. So are electric vehicles really more efficient than gas? Yes! They absolutely are, unless you have some very dirty electricity. So much for the well actually's.

 NewSection (8:34)

But climate change isn't just one problem. It's a hot mess of many problems at once. And it'll take some pretty radical changes from people all over the world to keep that hot mess from getting hotter.

 NewSection (8:47)

Glad we cleared that one up. Now, electric cars seem pretty intresting, but they still require that you drive them. And many see the future of driving as, well, not driving and letting a self-driving car do all the work for you. We've been talking about self-driving cars for a while now, though, and they have not become a thing yet. And I'm getting pretty impatient, so Stefan, please explain why I still have to drive. 

 NewSection (9:11)

By now you've probaby heard that self-driving cars are coming soon. If you haven't, suprise! They're coming soon! But people have been saying that for at least a decade. And I still can't buy a car that will drive me to work while I nap in the passenger seat. Some cars already come with partial autonomy. Systems like Tesla's autopilot that assists drivers or sometimes even take control. But they still need a human driver who can grab the reins on short notice if things get dicey. Which is why someone in the UK got arrested earlier this year for trying the passenger seat thing. There are some fully driverless vehicles that might be released in the next two years, but they're only meant for very specific uses, like long-haul trucking or taxis can find a certain streets and neighborhoods. That's because general-purpose driving is hard! The software has to work out a lot of really tricky question to turn information from its sensors into commands to the steering and pedals. And despite all the money and brainpower 

 NewSection (10:01)

 (10:00) to (12:00)

 that's being poured into research, there's still major challenges at every step along that path. The first thing a self-driving car has to do is figure out what's around it and where everything is. It's called the perception stage. Humans can do this at a glance, but a car needs a whole cornucopia of sensor data: cameras, radar, ultrasonic sensors, and lidar, which is basically detailed 3D radar that uses lasers instead of radio. Today's autonomous vehicles do pretty well at interpreting all that data to get a 3D digital model of their surroundings -- the lanes, cars, traffic lights, and so on. But it's not always easy to figure out what's what. For example, if lots of objects are close together, say in a big crowd of people, it's hard for the software to separate them. So to work properly in pedestrian-packed areas like major cities, the car might have to consider not just the current image, but the past few milliseconds of context, too. That way, it can group a smaller blob of points moving together into a distinct pedestrian about to step into the street. Also, some things are just inherently hard for computers to identify. A drifting plastic bag looks just as solid to the sensors as a heavier and more dangerous bag full of trash. That particular mix-up would just lead to unnecessary braking, but mistaken identities can be fatal. In a deadly Tesla crash in 2016,  the autopilot cameras mistook the side of a truck for a washed-out sky. You also need to make sure the system is dependable, even if there are surprises. If a camera goes haywire, for example, the car has to be able to fall back on overlapping sources of information. It also needs enough experience to learn about dead skunks, conference bikes, backhoes sliding off trucks, and all the other weird situations that might show up on the road. Academics often resort to running simulations in Grand Theft Auto. Yes, that Grand Theft Auto. Some companies have more sophisticated simulators, but even those are limited by the designers' imagination. So there's still some cases where perception is tricky. The really stubborn problems, though, come with the next stage: prediction.  It's not enough to know where the pedestrians and other drivers are right now,. The car has to predict where they're going next before it can move on to stage 3: planning its own moves. Sometimes prediction is straightforward: a car's right blinker NewSection (12:01)

 (12:00) to (14:00)

suggests it's about to merge right, that's where planning is easy, but sometimes computers just don't get their human overlord. Say an oncoming car slows down and flashes its light at you as you were waiting to turn left, it's probably safe to turn, but that's a subtle thing for a computer to realize. What makes prediction really complicated though, is that the safety of the turn isn't something you just recognize, it's a negotiation: if you edge forward like you're about to make the turn the other driver will react. So there's this feedback loop between prediction and planning. In fact, researchers have found that: when you're merging onto the highway, if you don't rely on other people to react to you, you might never be able to proceed safely, so if a self-driving car isn't assertive enough, it can get stuck: all actions seem too unsafe, and you have yourself what researchers call the "Freezing Robot Problem", which itself can be unsafe. There are two main ways programmers try to walk around all this: one option is to have the car think of everyone else's action as dependent on its own, but that can lead to overly aggressive behavior, which is also dangerous. People who drive that way are the ones who ends up to be the one swerving all over the highway, trying to weave between the cars; don't do that, by the way. Another option is to have the car predict everyone's actions collectively, treating itself as just one more car interacting like all the rest, and then do whatever fits the situation best, the problem with that approach is that you always have to over simplify things to decide quickly. Finding a better solution to prediction and planning is one of the biggest unsolved problems in autonomous driving, so between identifying what's around them and interpreting what other drivers will do, and figuring out how to respond: there are a lot of scenarios self-driving cars aren't totally prepared for yet, that doesn't mean driverless cars won't hit some roads soon. There are plenty of more straight forward situations where you just don't encounter these types of problems, but as for self-driving cars that can go anywhere, let's just say the engineers wont be out of a job anytime soon.

 NewSection (13:51)

Okay, fine. I guess I will wait just a little while longer. And while we're developing smarter cars, scientists can work on improving some of the other parts of driving, like the roads

 NewSection (14:00)

 (14:00) to (16:00)

 (16:00) to (18:00)

 (18:00) to (20:00)

 (20:00) to (20:27)