Brad Templeton Home

Main Page

Brad Ideas
(My Blog)

Robocar Blog

The Case



Car Design






End of Transit











Urban Planning






Sidebars: Charging

Speed Limits

NHTSA Levels

Valley of Danger


Google Cars

Random Notes

Random Notes

What about PRT?

Here is a sidebar on the issues of PRT vs. Robocars.

Reduced land for cars

It's been calculated that in extreme cases like Los Angeles, over half the land is devoted to the automobile; roads, highways, parking lots, garages, driveways -- it get's pretty big.

Over time, robocars should reduce the need for so much land. The biggest way they can do that is by reducing need for adjacent parking. In sprawl environments, stores try to have enough parking to handle the peak Christmas shopping days, but most of the time their lots are underutilized.

A robotaxi doesn't use parking at all -- it goes off and carries other people. But even an owned robocar that serves only its owner doesn't have to park too close. On peak shopping day, those vehicles could wait around in places like office and factory parking lots a short distance away, which are otherwise closed and empty. Of course, robocars can also park at valet parking density, and single-person robocars can pack 2-3 times denser than that.

Robocars can spend their time parked blocking driveways and fire hydrants. They can even double park or triple park along the roads. All the owner has to do is signal a need for their car via cell phone as they decide to walk to the door (or indoor electric robocar drop-off zone) of the mall, and it will be there waiting by the time they get there. (I know malls where your car could come from 5 miles away in the time it takes to leave them.)

And frankly, I suspect that the way robocars enable car sharing, there will just plain be fewer cars to store.

As this develops, cities will see another robocar dividend -- tons of land will become available for new purposes. Land in all sorts of useful places, for parks, offices, stores or housing. This in turn allows the cities to get denser, taking advantage of the greater traffic flow robocars should provide on the existing roads. Making trips shorter and more trips walkable to boot.


Dean Kamen hoped the Segway scooter would change cities. He didn't get his wish, but his scooter is quite interesting because of its small footprint and tiny turning radius -- it can be used to go a lot of places a traditional vehicle can't go.

However, it probably can't be a robot to drive people around, because it is important that you lean while standing on it. A tiny robot you stood on might throw you off if it turned or accelerated on its own, and you weren't ready.

The same would be true of a 2-wheel vehicle like a scooter or motorcycle. You need to lean and be part of the control process. But these are still useful vehicles. What robocar technology can do is deliver them to you.

A Segway might be more interesting if you could summon one quickly with your cell phone. Need to get somewhere fast where vehicle streets are not an option? You might be able to summon one. And when done, it would run off and charge and store itself. This makes it much more useful.

(I may be wrong about riding a 2-wheeled robot. After all, we are fully able to ride on the back of a motorcycle as a passenger, holding on to the driver. But we still have to pay attention and lean as it moves. Plus, I think it will be harder for people to trust the stability of such vehicles.)


As noted, a 2-wheeled vehicle may not be a suitable riding robot. At the least you would have to pay attention, unless it used fancy tricks to rotate your passenger compartment. (This might be worth trying, as 2-wheeled vehicles are very compact on the road and efficient.)

However, imagine a small robotic unit which attaches to a bicycle and, using the bike's wheels as well as its own, delivers a bicycle where you summon it. This could be a standard deliverbot that is able to hold a bicycle, but a tiny robot that uses the bike's own wheels could produce something super light and efficient, so bike delivery takes almost no energy, and doesn't need much battery. For stability, the combined bike and bot might have 3 or 4 wheels, like training wheels.


Service vehicles for the roads should soon become robotic. An interesting application would be the snowplow. Snowplows need to be heavy, but it should not be too hard to make fleets of plows which can immediately start roaming the streets in any snowstorm. Information could be broadcasts about what streets plows have cleared, so that robocars could change their routes to go to the most recently plowed areas. Most plows would find themselves followed by a long convoy of other vehicles.

With no need for staff, plows would go out as soon as there was reasonable work for them to do. It might be possible to design a system with more lightweight plows that move more often but only over areas of light snowfall, never letting the snow get so deep as to need a heavy plow.

It's also not out of the question that existing robocars, not in service for other purposes, could travel to a depot and have a plow put on their front, suitable for light plowing. When done they would return it to the depot. There might be much less call for heavy duty plows.

Robotic plows and snow-blowers could also zoom around cities clearing driveways and sidewalks, far more efficiently than the way it is done today. A network where the city (or private entities) could buy snow clearance in real time woudl allow all plows to be fully used during a storm, especially if combined with constant real-time readings of snow depth from the LIDAR sytems on the robocars.

There already is a robotic snow plow that uses magnets in the road. That's because in heavy snow, even humans can't figure out where the road is.

Unfortunately, today's LIDARs are not very good in very heavy snow. That might mean you want a human in the vehicle in such situations, but it need not be a full time plow operator, as most operations could still be done by the robot, which would still be able to do exact positioning if the LIDAR or advanced GPS can see anything, and know snow depth etc.

In addition, if the LIDAR detects there is a bump in the snow or something sticking out that could create a hazard, this could mean video is sent to a human being able to provide guidance.


What about expressing yourself, and your wealth, with your car? People like to arrive in flashy cars, and drive their friends around in flashy cars.

I suspect you'll still be able to spend a lot of money on your personal robocar, if that's how you want to show off. Car companies will always figure out a way to do that.

You'll even be able to spend more on your robotaxi, though this effort will be counterproductive if you end up having to wait a lot longer to get a rarer robotaxi rather than just taking the closest one. If you make an appointment (such as to take somebody on a date) you'll be able to impress. (Nobody is unimpressed if you come to pick them up in a limousine, after all, even though they know you're renting it.)

I suspect the future holds a lot of interesting display technologies which will allow people to put anything they want on both exterior and interior panels of a vehicle, as long as it's not too distracting to others on the road. So even if you rent a vehicle you might be able to load into it decorations which express your personality.

New York City

In the USA, NYC is the only city where most people don't own cars, thanks to lots of public transit and a giant taxi fleet -- and an inability of the roads and parking spaces to handle them.

New Yorkers who aren't taxi drivers should love robocars, which should come to them more reliably than taxis, move much faster and be safer and more pleasant than transit.

However, cities like New York do present a problem of "where to put them?" NYC is full of taxis, which keep in constant motion. To some extent they are "stored on the streets." There is much less on-street parking, and parking lot space is expensive and prized.

Single-person robocars may be the answer. These should be able to pack on roads 4 times more densely than ordinary cars, because they can go 2 to a lane, and can leave a shorter gap between them. They should be able to attain 4 times or more parking density too, because they will dense-pack, like valet parking. Finally, they can "park" in many places a car can't -- like low-use private driveways -- because they can be programmed to move on command. They might be allowed to park on the street in such driveways, and private owners might also rent out space to them. The private owners would be able to issue a command from computers or mobile phones that would make any robocars on their property temporarily scatter off it, and they could enter this command shortly before leaving, and their own robocar would issue it well before arriving.

The future of childhood

Studies show that 2 generations ago, children often ranged many miles from their homes. Today they are kept very close, rarely leaving the street and often driven to school. Robocars can offer unprecedented mobility to children, if their parents allow it. The location of childrens activities could vary greatly at no inconvenience to parents, once the children are old enough to ride alone in a robocar.

In some areas there could be fear (or overfear) of kidnapping keeping children away from solo riding. There may be a market for highly secure robotaxis able to safely move children. Once children reach the more responsible age (12) they will probably receive a newfound freedom.

Climate Control

When you get superefficient lightweight cars such as the single person vehicles I've been writing about, the energy budget for climate control becomes significant. It can easily take more energy to air condition a single person vehicle than it takes to drive it. To some extent there is no way around this, but there are some ideas that can mitigate the problem.

Because robocars and whistlecars can refuel themselves, alternate fuels for heating and cooling are possible. It may make sense to heat using small tanks of propane or even liquid fuels if the vehicle is otherwise electric. Just as short range lightweight vehicles don't need much battery for the trip, they don't need much fuel for the heating either.

Cooling is harder, but again range can be your friend. It's not out of the question that for a trip on a hot day -- and the robocar will know in advance of the trip just how hot it will be -- the car could simply load a chest with ice and use that for cooling. Or perhaps even liquid nitrogen. This would be totally impractical for a normal car, but a robocar could manage it by driving itsself up to an ice chute anywhere nearby before the trip and receiving a load of ice into its foam cooler.

Or one could just have bigger batteries and use traditional electric refrigeration. However, those batteries could come out in the winter, or quickly put back in if the forecast calls for a hot day. In dry climates, swamp cooling could also be considered -- it's much more energy efficient but only works in a more limited set of circumstances.

Of course, longer-range cars will tend to be powered by liquid fuels and combustion engines, just as today, and they can be heated and air conditioned just as they are now.

Robocar Reverse Pass

Here are some proposals for a different passing protocol when a slow robocar or robotruck is blocking traffic along a 2-lane (one each direction) road or highway.

If the vehicle is vacant, and not in a rush to get somewhere, it can and should pull off at any available turnout. In fact, the vehicles coming from behind might broadcast their location and speed on the motornet, so the slow/empty vehicle can pull off the road in advance, so the occupied faster vehicles never even notice it. However, this will not always be possible, nor is it fair for occupied vehicles.

A traditional pass is possible, of course, with the faster vehicle confirming it is safe to enter the oncoming lane and passing.

If 2 or more vehicles want to pass (as is common when several vehicles line up behind a slow vehicle) it is also possible to ask the slow vehicle to assist in a reverse pass. In this case, the slow vehicle would determine it was safe to enter the oncoming lane. It would do so, and immediately begin braking rather than accelerating. The faster vehicles would quickly pass it, possibly accelerating if need be. Once they were past, the slower vehicle would pull in behind them.

The reverse pass should move the other cars past the slow vehicle far more quickly than a traditional pass, and puts fewer vehicles into the oncoming lane. It involves disruption for the fewest vehicles as well. Because it happens more quickly, it is safer, as less time in spent by vehicles in the oncoming lane.

There is one danger. If many cars are behind the slower vehicle, or they are erratic, there is a risk that if a car appears in the oncoming lane, the slow car may have trouble returning to the regular lane. This would not happen with robocars, which would immediately brake to open a gap if the slow vehicle so much as starts swerving back into the lane. The unanswered questions involve unknown or untrusted vehicles, such as HDVs. One answer might be to require that the sixth car open a gap bofore the passing starts to let the slower vehicle know it only has to let 5 cars pass if it has to, and that this vehicle can be trusted. You don't have to trust all vehicles in the line, just perhaps one every four. Of course the law would demand opening a gap but we want more than the law here.

On multi-lane highways, passing on the right may also make the most sense if the lane is clear. Today passing on the right is technically illegal because it is an unsafe and unfamiliar pattern. This need not be the case for robocars.

Robocars vs. telepresence

While robocars may increase the range people travel, it will be interesting to see how trends in videoconferencing and telepresence alter how we meet. Videoconferencing is one of those technologies that has always been just out of reach, however there are more and more reports that now that bandwidth and video equipment have reached the level of doing multi-screen HDTV videoconferencing, there is something real there. Mostly people talk about using that to eliminate long airplane trips, but as the quality increases it may also eliminate shorter trips. It will be interesting to see how the two factors balance.


Small robocars, designed for just a few people, can be smaller and lighter than cars, and this means that roads and guideways for them can be much cheaper than full-use roads we build today. Astonishingly cheaper. Most residential blocks, for example, might be served by a single 10 foot wide lightly paved alley in the front or back. New construction might run such back alleys (as are common in some cities) with no roads at all.

The short roads could be one-way or rather "ad hoc" with computer coordination of direction and use. A few wide zones could allow coordinated passing on longer stretches.

However, this leaves the problem of trucks to move heavy items and to bring in the heavy equipment used in construction and rennovation. (Robotic fire trucks would be made smaller and broken up into sections, but travel as a fleet.)

Robotic trucks can be steered perfectly. That means you can have assurance on exactly where their wheels will go. This means you could build these light vehicle lanes but have two reinforced wheel tracks with much thicker asphalt. The robotic truck would never deviate from them. Alternately, the truck's tires could always go (slowly) in a shoulder area made of gravel or packed dirt. Even if a heavy truck does tear up the shoulder a bit, it's not too expensive or hard to manage. One might even imagine that road repair robots will exist to deal with this. No real rush on the repairs if they are needed.

This means residential streets that cost $100,000 to $200,000 per mile compared to the 2 to 3 million dollars per mile spent on typical urban streets.

The Disabled

The best way to serve the disabled with Robocars would be with special "shell" robocars that are empty inside, allowing the person to just wheel themselves in and ride in their chair. Such a robocar would have a small ramp and possibly be entered from the front or back to wheel right in. Electric cars are much more flexible about such things as they have no driveshafts and need not have ordinary axles.

This would provide tremendous mobility to the disabled. However, there would be fewer of these vehicles in the taxi fleets than ordinary robocars which come with a seat. This might mean slighly longer wait times as the vehicles must come from slightly further away. But as long as one is under a mile away, these times would be small, and trivial compared to todays paratransit wait times. Strangely, existing disability laws sometimes forbid "seperate but equal" solutions. I think it would be a great shame for both the disabled and the ordinary if some other solution which let the disabled ride in any robocar, at the cost of making every car larger and more expensive had to be chosen. The great thing about robocars is how they can deliver the right vehicle for every different need, including "special needs."


For best operation, robocars will have a data network to help them do their tasks. I have descriptions of what this Motornet might involve.

Next consider Robocar Oriented Development and the topics of urban geography. You may find the Glossary to be useful.