Tag Archives: driverless cars

Driverless Cars & Commute Times

One of many unknowns with driverless cars is what their impact on traffic will be. Boosters have predicted that driverless cars will reduce traffic by reducing the number of vehicles needed to serve travel demand and reducing following distances between cars. Detractors have predicted that these effects will be more than offset by latent demand for travel that will come out of the shadows when it gets easier to travel in a car, by being cheaper or less stressful, which will encourage people to travel more and longer distances.

We may be getting ahead of ourselves since we don’t know when true driverless cars will become available and achieve a large enough market share to have a big difference. However, it is worth considering what we can learn from differences in commuting time by travel mode.

The graph below presents the mean commute time for various travel modes for ten relatively large metropolitan statistical areas (MSAs) in the US. The data was compiled by Governing Magazine.

commute-mode-cities

Just about all cities have mean drive alone commute times of 25-30 minutes, with Portland being the low at 24 minutes and DC the high at 32 minutes. Transit commute times are about 1.5 to 1.8 times as long as drive alone commute times, and commuter rail commute times are about 2.0 to 2.5 times as long. Commute times are remarkably consistent across MSAs. Interestingly, Riverside has the longest commuter rail commute time, which reflects its dual role as its own MSA and as LA/OC’s more distant suburbs.

Since many transit riders are captive riders, especially in places like LA and Riverside, perhaps it is best to focus on the comparison between commuter rail and drive alone, since many commuter rail riders are choice riders. The data strongly suggests that people are willing to put up with much longer commutes when they don’t have to drive the vehicle themselves, which would support the idea that people will increase their commute length if they can use driverless cars.

This data, from a 2009 Census report, aligns with that conclusion. Note that carpoolers, who have to drive some of the time but not all the time, are willing to put up with longer commutes than drive alone, but not as long as public transportation.

census-commute

Driverless cars may be even more enticing than commuter rail, as there’s no inconvenience of having other passengers. On the other hand, unreliability caused by traffic may be a deterrent. The appeal of driverless cars will also be uneven, as certain types of work do not lend themselves very well to being done by oneself in a moving car. In addition, many people will not have the schedule flexibility to increase their commute; for example, if you have kids, you may not be able to or want to leave them in child care for a longer time every day. On the balance, though, it seems likely that if driverless cars become widespread, commuting times will get longer.

Advertisements

Trolley Problems

The New York Times has the latest in a long series of pop think pieces that wonder how driverless cars will deal with variations on the so-called “trolley problem”: is it ethical to make a decision that saves some people’s lives at the expense of another person’s life? This article asks whether your driverless car should hit a pedestrian to save your life. Not surprisingly, most people in the studies chose to save their own lives over that of a hypothetical pedestrian.

Should driverless cars be programmed to serve the greater good, even at the expense of the passenger?

How would you even know what the greater good is? What if the passenger has children in the car? What if the passenger is the only person in the car, but is the sole breadwinner for a large family and is taking care of disabled relatives? What if the passenger is 70 years old and the pedestrian is 35 years old? What if the passenger is 70 but in excellent health and the pedestrian is 35 with a terminal illness? What if the passenger in the car was on their way to commit a crime? Even if we, as a society, could agree on what the greater good might be, there would not be enough time to determine all the relevant information in time for anyone – human or machine – to make an ethical decision.

Quite simply, I do not think humanity is about to take the extraordinary step of allowing a fully automated system to decide who dies and who lives.

Instead, driverless cars will be expected to perform the same way we expect almost all other machines to perform: with an extreme deference to preserving human life. Think about the machines you interact with on a daily basis. You have to be negligent in the extreme to get killed by a machine because of something the machine couldn’t avoid doing (as opposed to something idiotic that the machine’s human operator might make it do). And even if you do something incredibly negligent, like step in front of a train or stick your arm inside rotating machinery, the machine’s owner might still end up with significant liability for your injury.

Many people view autonomous vehicles as incremental improvements to automobiles. And it’s true that there will be great improvements in safety from “driver assist” technologies like systems that help keep you in the lane and keep you from hitting the car or pedestrian in front of you. These technologies will save lives without a doubt.

However, full automation is not an incremental improvement. It’s a shift to a much different level of social/cultural expectations and liability. Drivers are held to extremely low levels of liability for damage they cause; in California, the state minimum is $15,000 for death to one person and $30,000 for death to multiple persons. In contrast, Metrolink paid out $4,200,000 for each death in the 2008 Chatsworth crash, and the number was only that low because of federal law that caps railroad liability at $200,000,000 per incident.

The reason railroads have much higher liability limits than drivers is that most people in the public identify as or with drivers, while very few people identify as railroads. If the state tried to raise the auto insurance minimums to $4 million per death, insurance premiums would skyrocket and there’d be a political revolt.

In other words, if you maim someone with your car, but you have the state minimum auto insurance and few assets, that person is shit out of luck. Google, on the other hand, is not going to have its liability capped at $15,000 per death. It has the financial wherewithal to pay for insurance that actually covers the damages caused by auto accidents, it has the assets to pay damages in excess of its insurance limits, and it’s not going to get any sympathy from the public if a driverless car runs over someone’s kid, someone’s mom, or someone’s grandpa.

It seems to me, then, that fully autonomous vehicles will by necessity take a very large discrete step towards eliminating deaths from automobiles. They will be programmed to do so by having very conservative software. The large corporations – and their insurers – responsible for the software, and maybe for owning and operating the vehicles as well, will demand it. No one is going to accept an incremental improvement in safety in exchange for a hundredfold increase in liability. Fully autonomous vehicles will only kill someone in cases where the victim is grossly negligent, and even then, there will likely be out of court settlements.

The nature of Silicon Valley frequently rewards entrepreneurs for being the first to the market with a product, even it means frequent incremental updates to fix bugs. As long as they don’t deal with the security of private data, software problems usually have minor consequences. Apps freeze and crash; Google Maps has its share of erroneous data; formatting in Word is still frustrating as hell. But as Theranos shows, other industries don’t work that way.

Railroad signaling may offer clues as to what will be expected of fully autonomous vehicles. Braking performance assumptions are very conservative. Automatic train control is not expected to be marginally safer than human drivers; it’s expected to completely eliminate train to train collisions. The system is designed to assume it’s not safe to move unless otherwise directed, not to assume that it’s okay to move unless informed otherwise. Railroads were just forced to spend billions on Positive Train Control, one function of which is to help protect railway workers against the trolley problem by insuring it never comes up in the first place.

It’s not that incremental improvements aren’t good. It’s just that cultural expectations change when we turn a task over to a machine. We don’t expect machines to make ethical decisions, we expect them to be safe enough that they never have to.

Driverless Cars and Driverless Trains

Updated with a note on platooning and some input from R Winston Kappesser (@ronaldkappesser).

There was some back and forth on Twitter today on the potential of driverless cars and their impact on rail infrastructure like transit and high speed rail. In that context, here’s a civil engineering perspective on the technological issues and potential impacts.

Stuck to You Like Rubber on Asphalt, or Steel on Steel, or Something

First, we have to understand the technological differences between rubber-tired vehicles and steel-wheeled ones. That starts where the rubber hits the road or where the steel hits the steel.

When it comes to transportation, friction is both our friend and our enemy. We need some friction; otherwise, when you hit the gas your tires would just spin in place, or your rail wheels would just do something like this. Friction between the rubber tires and the road’s asphalt or concrete surface is what keeps cars and buses from flying off the road at corners. It’s what turns the tractive effort of a big honkin’ locomotive into forward motion.

On the other hand, too much friction wears out your car’s tires and makes your car run less efficiently. For trains, friction management is a critical part of track and vehicle maintenance. If there’s too much friction, the rails and wheels will wear out faster, and the train will use more fuel, increasing maintenance and operating costs. Friction management is so important for railroads that locomotives are equipped with sanders, so that the engineer can drop sand on the rails to increase friction on upgrades, while sharp curves are equipped with greasers to reduce friction between the wheels and rails. There’s an entire sub-industry built around friction management.

In general, the coefficient of rolling resistance between rubber and asphalt is about an order of magnitude larger than that between steel and steel. This means there’s proportionally more rolling resistance between your car and the road than there is between an Expo Line train and the rails. The very low rolling resistance on railroads is part of why trains are so much more efficient at long-haul freight than trucks. The coefficient of friction is also lower for steel on steel than rubber on asphalt.

Can’t Stop, Won’t Stop

That efficiency comes with a cost, though, in braking performance. Trains can’t brake as fast as rubber-tired vehicles. How much worse? For a 70 mph design speed, Caltrans Highway Design Manual requires a stopping sight distance of 750 feet. For a 70 mph design speed in territory with cab signals, the standards used by Amtrak and many commuter railroads require a safe braking distance of 4,942 feet. For high-speed trains, the stopping distances for purposes of rail signal design can be in excess of 2 miles.

(Note: some vehicles, notably LRT vehicles and some high-speed trains, have electromagnetic track brakes that use electromagnets to “grab” the track, allowing the vehicle to stop much more quickly. These brakes are used for emergency only; safe braking distances for railroad signal design are calculated assuming no track brake is used.)

The other major difference between rubber-tired vehicles and steel-wheeled ones in this regard is the ability to steer. A person operating a rubber-tired vehicle has the ability to take evasive action to steer the vehicle away from a hazard, while a train engineer is obviously helpless to do anything other than brake.

Design Evolutions

These technological realities have resulted in a different evolutions of civil engineering design standards.

For cars, design is predicated on the driver being able to see further than the distance needed to stop the car. The design of vertical curves (changes between upgrades and downgrades) is governed by the need to ensure the ability of the driver to see over the top of the hill, or for the car’s headlights to illuminate enough of the road ahead of a sag curve. At horizontal curves, vegetation and other obstructions on side of the roadway must be cleared far enough back from the edge of the road to allow the driver to see around the curve. In a safe design for autos, the driver will always be able to see further than needed to stop the car.

In contrast, with the exception of low-speed streetcars, for trains it is simply impractical to design the track such that the engineer would always be able to see further than the distance needed to stop the train. Horizontal and vertical geometry of the track is controlled by vehicle performance and passenger comfort. Safety is ensured by the signal system providing the safe operating speed to the engineer, and in some cases enforcing that speed, based on the locations of other trains (or perhaps more accurately, based on information that sections of track ahead of the train are not already occupied by other trains).

Note the fundamental difference here. For cars, safety is based on the ability of the driver to passively gather information about conditions on the road. For trains, safety is based on active collection of information on the locations of trains, and active dissemination of instructions to trains that it is safe to proceed.

Driverless Technologies, and Others

This means that the interfaces and impacts of driverless technologies will be different for cars and trains. For cars, passive decentralized technologies (i.e. the car just gathers information, and doesn’t communicate with other cars or with a central control center) will suffice. For trains, centralized control is a necessity.

For cars, it will be a huge improvement for safety simply for driverless cars to more reliably and consistently do the things that we currently rely on human drivers to do. This will have some positive impact on practical capacity by reducing accidents. If driverless car technology allows cars to follow each other more closely than they do today, by eliminating the component of following distance related to human reaction time, that will increase road capacity.

For example, you may have already figured out that, despite the stopping sight distance being 750 feet at 70 mph, cars on a freeway flowing at 70 mph don’t actually space themselves 750 feet apart. At that rate, a freeway lane would only move about 500 cars per hour, but the actual capacity of a freeway lane is about 2,200 cars per hour. If you have a driver’s license, you may (hopefully) remember the “two second rule”, that you should leave about 2 seconds of travel distance between yourself and the car in front of you. At 70 mph, that’s a little over 200 feet – less than the stopping distance, and acceptable only because you can see further than just the car in front of you, and you have time to swerve out of the way if needed. Part of that 2 seconds is an allowance for human reaction time; if driverless cars allow that component to be eliminated, they will increase capacity.

On the other hand, safe design for trains is based on maintaining at least the stopping distance between following trains. At 70 mph, a train should never be less than 4,942 feet behind the train in front of it. In practice, the distance will always be larger due to the impact of grades and the use of fixed signal blocks. The engineer’s reaction time is portion of that stopping distance, but it’s not much. Driverless train technology has been around for decades, but the primary appeal is reducing labor costs, not increasing capacity.

If the goal is to increase capacity on rail transit, communications-based train control (CBTC) will probably offer more benefit than driverless technology, because it will eliminate the capacity waste caused by fixed signal blocks. CBTC should also allow railroads to take advantage of better braking performance available in newer rolling stock. The combination of CBTC and driverless trains would allow many transit systems to greatly improve service by increasing capacity and reducing labor costs, thereby allowing the agency to provide more service.

Lawyer Up

A big unanswered question, in my humble opinion, is the liability implications of driverless vehicle technology.

For cars, what will be the standard for safe following distance? At present, we allow drivers to follow each other at less than safe stopping distance. Will driverless cars follow the “two second rule” or will they be allowed to follow more closely? If there’s a rear-end collision, who is liable? Note that some of this must have been decided implicitly or explicitly by the people who have operational driverless cars, like Google.

For trains, at present, railroad signaling is based on the premise that the train in front of you is at a stop, and therefore you must be able to stop too. If you implement CBTC you could argue that if you know the position and the speed of the train in front of you (Heisenberg be damned), you should be allowed to follow more closely. On the other hand, if the train in front derails for some reason, it’s going to come to a stop very quickly, and any following trains that are less than the safe braking distance behind are hopelessly screwed. There’s not a consultant in business in the country today that’s going to sign off on allowing trains to follow each other at less than the safe braking distance, and I doubt any agencies would do it either.

Therefore, for trains, I really think the capacity improvements are going to come from CBTC, not driverless technology.

The Future is Uncertain

I hope this post doesn’t sound like it’s down on transit. For one thing, any improvements available to cars will be available to buses as well. And nothing is going to change the simple geometric advantages that transit enjoys in dense areas.

Predicting the future is hard. If you’re out there predicting the doom of the car every year like James Howard Kunstler or the dominance of self-driving cars in 2020 like Randal O’Toole, you’re probably going to end up looking foolish.

A conservative approach would be to continue investing in cost-effective transit improvements, including CBTC and driverless technologies, where warranted. Automated car technology should, at the very least, result in a considerable drop in the number of people killed and injured by cars, and for that alone, it should be welcomed.

Update: Platooning

A quick note on platooning, which is the idea of having driverless cars follow each other very closely, perhaps only inches apart. This would greatly increase road capacity, but it would absolutely require vehicle to vehicle, and perhaps vehicle to central control, communication, as opposed to passive information collection systems currently being used by the likes of Google. I think that getting such a system operate reliably and safely will be more difficult in practice than many people expect. I don’t think we’ll be seeing it any time soon.