How Driverless Cars Will Affect Transportation
What happens to transportation when self driving cars start hitting the streets? Many large corporations both in the automotive and tech space are in a race to become the first to offer a fully autonomous car. Some manufacturers are already offering certain aspects of automation, but no one has yet achieved (or been legally cleared for) Level 5, or fully autonomous driving. What can we expect when we begin to share the road with robots?
Goodbye Gridlock 🛑🚙
The first thing that is going to happen is that traffic congestion will decrease. Keep in mind that I’m just talking about the cars themselves. I’m not talking about the: no streetlights, every car on the road is a robot world depicted in sci-fi movies. Introducing self driving cars into existing city infrastructure has the potential to reduce congestion significantly. How? First, a primer. Check out the video below by CGP Grey.
Well that’s all fine and dandy when EVERY vehicle on the street is fully autonomous. What about the transition period (which could be multiple decades) where there will be a mix of autos and regular human operated vehicles? We will still be creating traffic snakes 🚦🐍 as long as humans are behind the wheels of any vehicle on the road.
How Many Self Driving Cars Do We Need? 🤖🚙
In the above video, the simple demonstration of a ring road showed how one car slowing down could cause congestion (🚦🐍) even long after that car is gone. The video also illustrated how just changing a lane on a highway can cause similar congestion under the same principle. There are already vehicles on the road today that have some automation, specifically when it comes to highway driving. Many cars can speed match the car in front of you, provide lane guidance to keep you in the middle of your lane and even automatic lane changing. If everyone had this in their vehicle, how much would traffic congestion be reduced? What if only a few people had it in their vehicles? Would it make any difference? This is the exact question that researchers at the University of Illinois - Urbana-Champaign set out to answer. It turns out, the answer is shockingly low.
"Our experiments show that with as few as 5 percent of vehicles being automated and carefully controlled, we can eliminate stop-and-go waves caused by human driving behavior” - Daniel B. Work (source)
Five percent. If five percent of the vehicles on the road are controlled automated vehicles, human caused congestion goes away. Now if you read the full paper (here’s a link if you’re interested) you’ll notice that the test environment was a ring road and was extremely controlled, so real world applications would be different. But, as we saw in the video above, it’s easy to take concepts from a ring road and see how they can affect every day driving. Think of the 5% self driving fleet as pace cars. These cars set the speed for the other 95% of (human) drivers and in doing so, eliminate human made congestion by controlling their speed. This is the most optimal situation, but it makes me curious about what happens when Level 5 autonomous vehicles begin to hit the streets of the real world. I don’t think we’ll eliminate congestion once we hit 5%, but we might come close once we hit 10% or 20%. The other thing to keep in mind is that this is just human-made congestion, “stop-and-go waves” as Work describes them (I still prefer 🚦🐍). What about congestion caused by volume or infrastructure?
Thank You Waze 📱🚙
Waze is a popular turn-by-turn navigation app that was acquired by Google in 2013. The killer feature that Waze perfected was automatic re-routing based on crowd-sourced real time traffic information. Whenever I need to drive anywhere, I check the route first in Waze. Even if it’s driving a route I’ve done hundreds of times, Waze can see traffic slow downs and accidents happen in real time and can automatically reroute me mid trip. Combine the self driving pace car fleet with real time traffic and congestion info and you start to get a multi prong attack to all types of congestion, not just the human made kinds.
What Happens to Ride Sharing? 🚕📲
Some of the biggest players in the race to create the first self driving car are the ride sharing apps. Uber, Lyft, Didi and others are investing significant resources into developing their own self driving cars to replace their human drivers. While the ride sharing companies have routinely said that their services eliminate grid lock and traffic congestion, a new study shows the opposite.
Between 2010 and 2016, weekday vehicle hours of delay increased by 62% compared to 22% in a counterfactual 2016 scenario without [ride sharing companies]
In addition to this data, ride sharing operators are highly motivated to move to self driving vehicles instead of the human drivers they have now. Not only will they eliminate their biggest expense (paying the driver), but they can much more efficiently control supply and demand. Big conference coming to town? Uber can program some of their self driving fleet to change cities to accommodate the increase in demand. These cars bring themselves to the conference area and go home when attendees leave. Uber apparently has an order of 100,000 self driving cars from Mercedes set to deliver next year and Lyft allows users to order Waymo self driving cars from their app right now in Phoenix. These ride sharing providers are headed towards a self driving car future, and that will have a significant impact on the professional driving industry.
Conclusion
Self driving cars are coming. It is a fact. Transportation industries around the world will have to adapt to this huge paradigm shift. Trucking, shipping, taxis, ride sharing (and more) are all going to have massive shifts in the way they operate. Many are focusing on how these new technologies can liberate us from gridlock and aging infrastructure, myself included, but there are other questions. How will society accept (or not?) self driving cars in practice? What problems will they bring with them, and how will we overcome those challenges?