Autonomous-Ways, or A-Ways, Part 1 – What Do We Want?

When I focused on how Autonomous Vehicles could transform transportation in 2007, I started by looking at my own transportation habits and needs. 

I rapidly realized most travel is local, so improving local transportation is the key, not bullet trains and ultra-fast long distance services (although, I’ll address those as well). My typical trip around town is on roads with speed limits of 25 mph, 35 mph, and 45 mph. Thus it takes 14 minutes to go the 6.8 miles to the Garden State Parkway, for an average of 29 mph. Traffic adds about 5 minutes, which brings the average down to 23 mph.

I spent a lot of time contemplating our roads and highways, usually while stopped at stoplights and stop signs, or stuck in slow traffic – there are 8 traffic lights and 2 stop signs on that trip to the Parkway.

Simple things like eliminating stop signs, traffic lights, and slow speed limits could almost cut my typical trip in half. I quickly realized that Autonomous Vehicles don’t need such traffic restrictions because inter-vehicle communications, using my ideas for Cloudlet Computing & Communications and traffic management would keep them from running into each other.

With this careful management Autonomous Vehicles can drive closer to each other, both front-to-back and side-to-side. This means we can get a lot more cars into the same roadway area: today cars occupy only 11% of a roadway. This will reduce congestion – which speeds up my typical trip even more.

Close coordination also means Autonomous Vehicles can safely travel faster on the same roads. Now we might do my 6.8 mile trip in 6.8 minutes, or even less. 

I call the “roads” that these Autonomous Vehicles will move on Autonomous-Ways, or A-Ways J

An unintended consequence of making transportation faster and more pleasant is that people will travel more. So any efficiency from not stopping and starting would be gobbled up by the additional travel.

That got me thinking about how to make Autonomous Vehicles much more efficient than cars. Cars are very inefficient, so that isn’t too hard. Overall transportation is about 20% energy efficient, and most of the energy from our imported oil goes out the tailpipe of our transportation system.

Autonomous Vehicles won’t have collisions. If we don’t have any collisions then we can eliminate a lot of weight from current cars: bumpers, crumple zones, super strong doors and frames, air bags, etc. This makes Autonomous Vehicles substantially lighter than cars.

Weather is another hazard, both to driving, and to the roads themselves – during freeze-thaw cycles here in the Adirondacks you can almost watch the roads disintegrate. This led me to think we need to enclose these new segregated ways.

Once we are free of weather related issues, and the vehicles are lighter, the enclosed ways can be engineered to be very flat and smooth. This means Autonomous Vehicles don’t need heavy tires and suspensions, which makes the vehicles even lighter and smaller. This greatly reduces rolling resistance 4%-8% of losses for current vehicles.

As I discussed in Optimized AutonomousVehicles, I believe we could get the Autonomous Vehicle replacement for car-like functions down to less than the weight of the contents they are carrying – more like an enclosed electric bicycle.

Of course there will be a wide variety of vehicles, from the tiny “pill-vehicles”, to Personal Mobility Vehicles, to Convoy Vehicles that carry many people and large quantities of other items, including other Autonomous Vehicles nested inside. So we aren’t really talking about a single A-Way, but a whole family of them. Here I’m focusing on carrying people over the distances they typically travel in a car, a few miles. (I’ll deal with very short distance and long distance A-Ways later.)

Once the Autonomous Vehicles are close to each other, why not actually link up, like train cars. This allows us to greatly reduce air resistance (4% for local travel, and up to 16% of highway losses for current vehicles), and air resistance losses grow as the square of speed, as we go to 100 mph, and even 400 mph, this will become even more important.

Originally I imagined the individual Autonomous Vehicles just linking together to form the Convoys. But I soon realized that to get the higher speeds we want, safely and efficiently, we needed longer, multi-passenger vehicles.

Now we have separated the vehicles into high speed Convoys and lower speed Personal Mobility Vehicles. And I noted that the Personal Mobility Vehicles could carry other items, up to the weight of the person. That led me to the concept of separating the Content Carrier from the Mobility Platform, so that you had a seat and the associated amenities when carrying a person, but could optimize the Carrier for other items, whether refrigerated, or heated, or fragile, and allowing for nesting smaller Autonomous Vehicles.

Once we link them up, that led me to the concepts of Continuous Convoys, Nesting Autonomous Vehicles and En Route Sequencing, which allows the Personal Mobility Vehicles to move within the Convoy Vehicles, and brings us to an integrated Transportation System, so you can change from one mode to another without getting up.

But to achieve those gains we have to keep Autonomous Vehicles away from us, dangerous drivers that we are. Drivers and cars caused 35,900 deaths, 2.2 million injuries, and 4.3 million collisions with “only” property damage in 2009; 1/3 of deaths involved speeding, 1/3 involve driver intoxication, and 22% of injuries involved driver distraction. Between 91% and 99% of crashes involve driver behavior issues.

And of course we have to separate those delightfully speeding Autonomous Vehicles from pedestrians, bicyclists, kids, pets, and other critters. This started me on the path of envisioning separate “ways” for the Autonomous Vehicles.

But where to put them: obvious places are on medians of current highways, or as an “Autonomous Vehicle only lane”, or supported above a current highway.

How do you get to those Autonomous Vehicle-only lanes, and how do you keep drivers from going into the lanes. I contemplated all sorts of fancy schemes.

Intersections pose major challenges – even if you have separate Autonomous Vehicle-only lanes, how do you get across the conventional roads at the intersections. Bridges block going above some highways, and the bridge supports often block the median.

Bury the Autonomous Vehicle-only lanes? Expensive and you still have the problem of bridge supports, not to mention all the utilities running alongside existing roads: utility poles for power and communications, buried water and sewer lines, buried communications lines.

What about making all the roads Autonomous Vehicle-only? Where could we walk, or bike, or move things that didn’t fit in Autonomous Vehicles, like construction vehicles and building materials.

This seemed like a dead end because we couldn’t make the transition from our current, obsolete, infrastructure to a new paradigm – I needed better ideas.

In the next Post I'll describe my vision for Autonomous-Ways.

Continuous Convoys & En Route Sequencing

One of the major tradeoffs in high speed mass transportation systems is the speed of an express vs. the convenience of the local with many intermediate stops.

My innovation of Continuous Convoys & En Route Sequencing give us both: each person receives non-stop service from where they get on to where they get off.

The basic idea of a Continuous Convoy is simple: a Convoy goes racing along at full speed. As the Convoy approaches a station, the passengers and items boarding at that station are already aboard a Convoy Vehicle, which accelerates and let’s the Convoy catch up, so it is at the front of the Convoy. As the Convoy approaches a station, the Convoy Vehicle at the end of the Convoy has all the passengers and items for the next station, detaches prior to the next station, decelerates and stops at the station. Passengers and items leave the vehicle, and the cycle continues.

The basic idea of En Route Sorting is also simple: after your Convoy Vehicle attaches at the front of the Convoy, you need to move to the Convoy Vehicle which will stop at your destination station.

Things get a little complicated if the Convoy Vehicle that will stop at your station hasn’t joined the Convoy yet; then you will need to move while the Convoy is in motion to get to your desired Convoy Vehicle.

This is where it helps if you are already riding an Autonomous Vehicle which is riding Nested in the Convoy Vehicle, because then your Autonomous Vehicle can take care of moving you to the proper Convoy Vehicle without your having to move or even think about it. Thing get even better if you need to change from one Convoy to another, because your Autonomous Vehicle an take care of that transfer for you too.

Sorting things en route is an old idea, for example, from 1864-1977, the Railway Mail Service used clerks on board moving trains to sort and process mail, speeding delivery.

You get the full benefit of the speed of the Convoy, with only one acceleration and deceleration at the beginning and end. This also saves a lot of energy, because only the Convoy Vehicle with the people and items destined for a station stops there. Because the kinetic energy is proportional to the square of the velocity (E=Mv²), this energy becomes important the higher the speed of the Convoy. Another benefit is that the energy regained with regenerative braking from the Convoy Vehicle as it approaches a station can be transmitted through the electric feed to power the acceleration of the nearby Convoy Vehicle just leaving the station.

This approach can be used with existing trains if the cars are individually powered. So the ideas can have immediate benefits. For existing trains, Intercity rail travel is 27% more efficient than Commuter rail travel (2,271 vs. 2,897 BTU/passenger mile, http://cta.ornl.gov/data/chapter2.shtml table 2.12).

Note the “station” does not have to be stationary – it might be another Autonomous Vehicle connecting to another route or to another mode of transportation (including faster or slower). This allows moving from one Convoy to another without stopping, and thus presages a network of Autonomous-Ways for Autonomous Vehicles, as described under A-Ways.

For a detailed example, including graphics of passengers and Convoy Vehicles, see Continuous Convoys and EnRoute Sequencing, How Many Vehicles in a Continuous Convoy?. Another analysis is in Continuous Convoys for NYC Subways.

The example shown there is for travel by the existing railroad from Long Branch, NJ to New York City. 

Here are some further results of the time savings possible with Continuous Convoys and En Route Sequencing for that route. The calculations show it is faster to drive than to take even the Express  train (10 stops). But even at only 60 mph, the Convoy trip is 13 minutes faster that a car (and that’s assuming no traffic congestion, lots of luck), and 25 minutes faster than the Express.

Ironically, making the train go faster doesn’t help that much because it spends so much time loading, unloading, accelerating and decelerating that it doesn’t gain much advantage from the theoretical maximum speed. The Continuous Convoy does take full advantage of the speed, so an Acela-like speed of 120 mph gets your there in 30 minutes, and with a bullet train speed of 240 mph, you are there in 16 minutes, vs. 55 minutes for the Express.

Thus I believe Continuous Convoys and En Route Sequencing can have near-term results in getting people out of inefficient cars and into mass transit.

The train trip distance from Long Branch to New York City is 55 miles, with 10 or 18 intermediate stops for the Train, and no intermediate stops for the Convoy. Driving time is per Google Maps, and traffic conditions will add to the time.  I’ve included different maximum speeds for the Train and the Convoy: note in the 240 mph case, the Train can only accelerate to 200 mph before it has to slow for the next stop, assuming the stops are evenly spaced; including all 18 intermediate stops, the Train reaches a maximum speed of only 155 mph. For those interested in the assumptions:  0.1 g acceleration and 120 second stop at each station to load and unload.

Another analysis is in Continuous Convoys for NYC Subways.

I spent many days trying to think of how to use existing tracks and stations, that is the existing infrastructure, to achieve a much faster commute. Finally I came up with Continuous Convoys and En Route Sequencing.

Later I’ll present additional ideas to deal with the realities of the Long Branch-to-NYC train: the tracks aren’t up to the higher speeds, there are a lot of at-grade crossings, and other trains use the same tracks.

But you can apply the same ideas to Amtrak trains to achieve similar gains. For example, I wanted to take the Acela from Metro Park, NJ station to the New Carrolton, MD station, but none made both stops. Of course we need new vehicles: each Vehicle needs to operate autonomously, with its own power, so we still have some work to do. More on that later.

Next we'll talk about Autonomous-Ways or A-Ways.