Dualmode and Automated Highway Systems (AHS)

by Brian Dudson

I joined the transit-alternatives discussion group some weeks ago and have been hanging in, trying to catch up. I have checked Jerry Schneider's exceedingly helpful page 'Supported automated electric advanced transit technologies' and looked at most of his references to various PRT schemes. Then, I found the reference to the Dualmode Discussion Group. I don't normally lavish superlatives, but the more I read the better it got. It's almost restored my faith in human nature as I realized that there is a future for urban transport. I've summarized my understanding of DM possibilities, which, though far from complete, enables me to put my own views into context. I hope to contribute by offering a another perspective.

My name is Brian Dudson and I am a supposedly retired architect/city planner. I was born in New Zealand but have lived in Brisbane, Australia, for over twenty years. I studied architecture in Auckland and architecture and city planning at the University of Pennsylvania. I have worked in Auckland, Hong Kong, Philadelphia, Washington D. C., Malaysia and Brisbane.

My interest in transport goes back almost to my beginning. My mother said cars were the first thing I drew and as a kid, about first grade, a friend and I played 'roads' on the bank of a stream. It was a clayish sort of sand which cut and held an edge quite well. I made the roads, tunnels, bridges and garage burrows. My friend trundled his toy trucks around, then grew up to own a trucking business. I hate to think of the damage my parents could have wrought if the had bought me a train set, not that the family finances could have run to much.

I am not a typical city planner of today, as I believe planners should at least try to plan for what people want and, as even the dimmest unemployable knows, people want cars. So I wrote a book To the Cities of the Second Car Age, which argues that cars will become clean, sustainable and automated. Automation will make cars safer and increase the capacities of highways to such an extent that, theoretically, all motorized trips in even the largest cities could be made by car - and this in a way which calmed surface streets to the extent deemed desirable.

My early reading on AHS included a 1982 report prepared for the USDOT and GM called 'Systems Studies of Automated Highway Systems'. Not an easy read, but it described preliminary, yet remarkably comprehensive investigations, into the feasibility of automating highways in and between cities. It choose Philadelphia as a basis for these investigations.

The systems it considered included: (1) most of the intelligence on-board vehicles; (2) most intelligence wayside; (3) detachable electronic packages; (4) pallets: and (5) partial pallets, in which the front wheels rode a small control pallet while the rear wheels continued to propel the vehicle. (There were few front wheel drives at that time.) The report also considered the feasibility of electric vehicles which, because of their precise tracking, picked up current from the guideway by conduction or induction.

Its full pallet system was not unlike some of the DM systems now being discussed. So what is PRT and what is AHS? I'm not familiar with a definition of PRT but it would have to be something like the automated movement of vehicles along segregated guideways. AHS would meet these criteria, however Flexitrain and Autobus, which are discussed by the group, would not (not that it matters.). So AHS can be seen as PRT and I feel it is a pity that it doesn't have a contributing contender in the group.

(Although various forms of AHS have been mentioned, from here on I refer to AHS as essentially the type of system developed for the National Automated Highway System Consortium (NAHSC), as demonstrated in San Diego in 1997, and now taken up by the group called the Cooperative Vehicle-Highway Automation Systems (CVHAS).)

AHS was de-funded because neither safety nor capacity would be improved if automated vehicles shared lanes with conventional vehicles, and because highway authorities considered separate facilities were neither financially nor politically viable. There were also serious concerns over the mismanagement of NAHSC.

I have the view that one of the more desirable systems would fall somewhere in the gap between AHS and DM. In this respect my thoughts correspond with those of Joe Palen, which he expressed in the first posting to appear on the DM Debate page (February 1997). This is discussed later in connection with safety.

So what's in a name? Would a rose be as sweet, etc? As the current reverence for light rail shows, perceptions can be more important than reality. It's not what Abraham Lincoln told us, but it's possible to fool enough people long enough. As we are aware, most politicians and planners calling the shots seem to be gung-ho for public transport and opposed to highways. So a system called PRT is more likely to get underway than the same system labeled AHS. But you can't fool everyone forever and what really matters is which system, or combination of systems, will provide the best feasible form of urban transport for the future.

What will people want?

The people of free countries will themselves continue to determine what transport they prefer. Future populations will be more affluent and better able to pay for what they value. Of course future transport will not result merely as the sum total of individual choices in the market place, as societial concerns must also be met to satisfy aspects such as pollution and sustainability. (I say little about these aspects in this submission as it is already long enough. However, I believe they can be addressed satisfactorily for all of the systems being considered, including AHS.)

It is improbable that people will accept anything less than a door-to-door, no wait, no transfer system. Cars normally provide this and the fact that cars now provide about 98% of motorized trips in American cities must mean something. So must the fact that, despite Europe's densely developed cities with their growth directed along transit corridors and, despite punitive taxes on cars and petrol, the English are driving almost as far each year as Americans, and Germans and Italians now own more cars per person (although this is partly because of a growing preference of many Americans for light trucks). As DM proponents are aware, future transport will have to be like improved cars - cars without congestion - cars which park themselves.

I can't see single mode PRT as being much more than improved public transport, nor much more popular. It would be more appropriately termed 'personalized public rapid transport'. It could only be popular if it were perfect, with guideways along every street and vehicles waiting close to everyone's front door. But as Kirston Henderson points out, such an extensive system of elevated guideways, specifically designed to keep costs to a minimum, would still be prohibitively expensive. Perfect PRT is almost as improbable as perfect public transport, which would require something like busses or trains along every street every few minutes. (Actually, perfect public transport could well be possible one day when all streets in a city are automated, as it could then be provided by public automated cars.)

The size of vehicles will remain important. The very small, limited range vehicles currently proposed for many DM systems, may have an initial fascination, but cuteness palls and it is unlikely they would remain popular. (Freeways were exciting at their outset.) I can remember vehicles of a similar size, such as the Gogglimobile, Isseta and the two seat tandem Messerschmitt, which were produced for the best of reasons. They came on the market but soon disappeared. It must also be relevant that, whereas American cars were giants amongst their counterparts from Europe and Japan in the 1950s, this is no longer the case. American cars lost their fins and excessive overhangs and European and Japanese models got progressively larger. I believe the range of sizes of today's cars is about optimum to satisfy consumer choice.

DM will have to operate in competition with freeways, possibly in competition with automated highways. DM will be attractive if it avoids congestion, but this attraction will be limited if it disappoints in other ways, such as being overly expensive to use. Or, if it requires small, not so comfortable cars of limited range, which people must purchase in addition to their conventional cars. Then people will be pleased to see others use a DM system so they can continue to drive on less congested freeways. Clearly DM can no more than supplement freeways at their beginning, but if DM provides pretty much all of what people want, then the normal operation of the market will see it replace freeways.

Urban network  

Critics of DM have claimed that it will merely transfer congestion from freeways to surface streets and that it will not be possible to park all of the DM vehicles attracted to already congested parts of a city, such as the central area. This need not be the case if guideways are built along the principal arterials and into the centers of cities, and if additional, readily accessible, parking is provided.

The ability of automated vehicles to park themselves is a major advantage. Spurs can be constructed from guideways into a CBD, for example, so cars can drop their passengers at a central station, then take and park themselves in some out of the way location until they are summoned. How this would enable all motorized trips to access the centers of large cities is calculated and illustrated in my book. I'm not claiming this as an objective in itself, and there will remain a need for public transport and service vehicles. The objective is to provide the transport people want, and I have illustrated my vision of a likely outcome of meeting this objective.

It was calculated that to provide all motorized trips to the center of Los Angeles or Sydney by car would require 18 automated lanes and parking equivalent to two structures covering a city block and 30 levels high. All motorized trips to the center of London or Paris by car would require 40 lanes and 44 such parking structures (the illustration shows five levels of four lanes on either side of the CBD). Similarly, the center of New York or Tokyo would require 72 lanes and 88 such parking structures. (The 72 lanes could be provided in tunnels equivalent to 12 conventional two lane tunnels.) Any or all of the automated facilities could be provided above, at, or below ground level. (See Figures 1 and 2).  The drawing of a London or Paris type central area shows about a quarter of the CBD, more or less to scale.)

The estimates were based on automation increasing the capacity of highway lanes three times. Obviously, if a DM system could increase the capacity twelve times, the number of lanes would be a quarter of the number mentioned above.

It will be necessary for all DM systems, including AHS, to have their own segregated guideways and entry and exit ramps. Guideways are likely to be located along freeways, as freeways were located to satisfy trip demands, and because their rights-of-way exist. DM networks must be like highway networks, with interchanges connecting all branches, rather than like subway rail systems, where lines cross but seldom connect, thus requiring passengers to transfer. Not all cities have completed freeway networks and some have none at all. Already congested arterials will become more congested unless guideways are provided to supplement their flow.

(A two lane busway is nearing completion along a freeway close to us, complete with its own on and off ramps and stations. It's costing half a billion dollars but is unlikely to increase bus patronage to any marked extent. There was almost no opposition as, after all, it is public transport. I mention this busway to indicate that it is not necessarily politically impossible to build new facilities along existing freeways. I like to think the busway can be converted to something more useful in the future.)

Contending technologies

Capacity, headway control

There is no point to DM systems unless they achieve significantly higher capacities than conventional freeways, which now carry, in rounded figures,  about 2000 vph per lane. Primarily, capacities must be increased by the use of electronic sensing and/or control  to greatly reduce headways between vehicles.

Prior to reading the Dualmode Debate Page, I understood that, possibly subject to patent rights, electronic sensing and control technologies would be essentially the same for all systems. If they were running at the same speed with vehicles of a similar size, their capacities would be about the same. Thus, I thought the disparity in the capacities claimed for various systems resulted, to put it politely, from varying degrees of optimism of their promoters.

I also thought the very high capacities mentioned would require pre-platooning. I now believe (and hope) this is not the case, as pre-platooning greatly increases the complexity of a system's operation. (Say a mature guideway network had 50 exits, then, theoretically, it's necessary to form 50 platoons at each entry. Possibly the number could be reduced to something like the number of branches in the network, if platoons were formed on a last-on, first-off, basis, but this would introduce other complications.)

I now understand that the headway control for systems using linear synchronous motors (LSM), such as InTranSys and HiLoMag, and the less developed proposals of Kim Goltermann, are quite different. Positioning is controlled by the system's computers rather than vehicle intelligence and radar or laser type sensing. The cycles of the alternating current allow perfect position data and accurate dead reckoning, so merges are perfectly timed. A LSM guideway has been likened to a conveyor belt of boxes (empty or loaded) though they are non-physical, electronically defined boxes. I visualize the operation as something like running lights, as for Christmas decorations, where the vehicles follow the pulse. A LSM guideway operates at constant speed, with all acceleration  and deceleration occurring on entry and exit ramps.

The estimated capacity for a guideway incorporating LSM at 60 mph, is said to be 12 times that of a freeway lane, or 40 times between cities at 200 mph, an enormous advantage. By comparison, the capacity of MegaRail, which would operate at 70 mph in urban areas, is about 3.5 times greater than that of a freeway lane. The capacity of an AHS lane (presumably at 60 mph) would be 2 to 3 times that for a conventional freeway lane. This is still a worthwhile improvement. William Turnbull has pointed out that the peak one-way flow on Los Angeles freeways is usually between 6000 to 10000 vph, although one achieves 12000 vph.

Of course it is the capacity of the whole system which is relevant. Critical to this is the capacity of its switches. There seems to be a consensus that switches which rely on the movement of the guideway, like railway points, are far too slow, and that passive or on-board type switches are essential. It is claimed that most of the contending systems have such switches and guideway capacities would be maintained, although so far, only AHS has demonstrated this ability. Kim Goltermann, a LSM advocate, says switching is the biggest technological challenge for DM and that switches may define the design of a guideway.

Palle Jensen says that because RUF vehicles would be closely coupled, RUF would achieve capacities 30 to 50 percent greater than AHS. However, this capacity would not be maintained at its switches, where vehicles must slow to 20 mph, come off the guideway they normally straddle and negotiate the switches on their street wheels. The resultant reduction in capacity (2000 to 3000 vph?) is a major limitation.

This reduction in speed does, however, allow bends in guideways or ramps to be made at much reduced radii, thus making them easier to fit to the fabric of a city. Certainly guideways can be angled to allow cornering at high speed, but there is a limit to the Gs a granny can pull. (My proposals split high speed lanes into two low speed lanes where they turned into parking spurs.)

Although it is not a problem of immediate concern, another aspect needs to be considered if DM systems are to eventually replace freeways. While it will be relatively easy to construct additional guideways to meet demand, it would be far from easy to provide separate entry and exit ramps for each guideway. On and off ramps need to serve multiple lanes, with vehicles merging between lanes to and from them, as on conventional freeways. This should be possible in any systems where the capacity of the switches matches that of its guideway. Also, this should be less of a problem for LSM systems, as they would have less need for multiple lanes.


All of the DM proposals have elevated guideways in order to facilitate their location in built-up areas. Some have vehicles running on top of their guideway, others have them suspended below. The guideway can be monorail, dual rail, or in the case of AHS, a road. (Note that AHS could readily operate on an elevated dual rail structure, something like narrow MegaRail guideway, which could also be relatively light weight if it only carried cars.)

Most DM proponents expect opposition to elevated guideways in some residential areas, but, speaking as an urban designer who spent more time than he has cared to working with representatives of different community groups and organizations, I don't think the extent of the virgorous opposition likely to be encountered is fully appreciated. It will be equally important to locate guideways at grade or underground, something which is easier if vehicles run above their guideway. It's not that I relish the idea of people moving about their city like rats in a sewer. They should be above ground enjoying the view, but trade-offs have to be made and tunnels will be necessary in the more precious parts of cities. Numbys are not as vehement as Nimbys.

Cost should not be a problem. Six car-only lanes can be provided in a tunnel having the same cross-sectional area required for two conventional lanes, and because of the absence of trucks, the capacity is increased fourfold. Such a car-only tunnel is now being constructed under Versailles as part of the Paris A86 orbital expressway. If automation triples the capacity of each lane, the capacity of the tunnel is increased 12 times. On a vehicle/mile basis, the cost is about half that for a typical suburban freeway, about the most cost-effective transport facility of today. Increase the capacity of each lane 12 times, then we are way, way, ahead. All DM systems should be promoted on the basis of cost per vehicle/mile, rather than mile.

Other contentions

Much of the material on the Dualmode Debate Page concerns the relative advantages or disadvantages of various means of:

1. propulsion (electric rotary motors, LSM, internal combustion engines (or whatever evolves to propel vehicles on roads);

2. traction (conventional or improved pneumatic tires, steel wheels on rails , maglev)

3. lateral guidance (horizontal or angled wheels along the sides of a rail, maglev, or magnetic plugs in a pavement.

My lack of expertise on these matters is such as to justify only the most general comments. As previously mentioned, the superior capacities, control and safety  claimed for LSM systems would make them orders of magnitude superior. But they could take longer to develop and they would cost more, if for no other reason than it would be necessary for their vehicles to have two mean of propulsion, one for guideways, one for streets. Systems incorporating electronic sensing for headway control offer real advantages in the shorter term.

The differences in the performance of the other technologies referred to, do not appear to result in such significant differences in the level-of-service they provide. Thus, my attitude is essentially 'whatever'. Maglev may well be essential for the very high speed inter-city systems, but not within cities. However, like LSM, its braking depends on electro-magnetic forces rather than friction, so it should not be affected by weather. Also, it is said that both maglev and LSM can be provided at about the same price.


Pallet systems have the great advantage of enabling conventional cars to use a DM system at its outset. however, they add to its cost and operational complexity. There would need to be a considerable amount of dead-heading, particularly at peak hours, and there would need to be a reserve of pallets at each guideway entry to meet the 'no wait' criterion. Francis Reynolds says half of the pallets in some flows would be empty. Calculations made by Kim Goltermann indicate that one pallet would be required for ever two vehicles using the system and the cost of a pallet system would be twice that of a true DM system. William Turnbull disputes this and Kirston Henderson says these problems can be handled.

Those who oppose pallets point to the time and facilities needed to load and unload them, one estimate being that 10 loading bays would be required to achieve 3600 vph. However, Kirston Henderson explains that vehicles would not need to stop to use MegaRail's 'ferries' and its ramps would achieve flows equivalent to those of freeway ramps.

Palleted systems do not readily lend themselves to automated parking, as either the pallet must stay with the vehicle or transfer its charge to some other sort of pallet. Another major disadvantage.

Pallets put all operational parts of system under the control of a systems operator, so regular inspections and maintenance should be a matter of course. However, those who favor true DM say private vehicles would be checked electronically, virtually instantaneously, each time they tried to use a guideway, and that owners would be required to have vehicles regularly maintained.


It is generally envisaged that guideways will be owned and operated by some form of systems company or authority, whereas DM vehicles would be privately-owned like cars today. A number of implementation strategies have been discussed concerning financing and overcoming the chicken-and-egg hiatus.

It is highly desirable for systems to be designed, like MegaRail and HiLoMag, to take both DM vehicles and pallets. The pallets would enable conventional vehicles to use the system as soon as it opened, but would be largely phased out as people purchased more DM vehicles. User charges would need to be high at the outset to cover the cost of the pallets and their operation, nevertheless there should be many pleased to pay to avoid freeway congestion.

Another approach, favored by both MegaRail and RUF, is to introduce their systems as public transport, on the basis that political support and funding is more readily available for public transport. In addition to its personal DM vehicles, RUF provides mini-buses which can link up as trains while on the guideway. There has been debate on the expense and logistic problems involved with the bus drivers.

The third strategy is to design a system based on minimal changes to conventional vehicles to enable them to operate on automated guideways. This approach is being adopted by CVHAS and is also the stated policy for introducing AHS in Japan. The necessary sensing and control intelligence will be progressively added to cars as adaptive cruise control, collision warning systems, then collision avoidance systems, etc, so that in a few years, enough vehicles will have enough intelligence to justify the provision of automated guideways. I have been advised that once production is underway, automation could add about 10% to the cost of cars and 10% to the cost of highways. This should be largely off-set by reduced accident costs. As others have mentioned, it is possible manufacturers will somehow absorb the additional vehicle costs over time, as they have done with features such as airbags.

Kim Goltermann also favors minimal changes to conventional vehicles, though he would have manufacturers bolt on a few magnets so vehicles could be propelled and controlled by LSM on the guideway. Maybe it's that simple. However, Francis Reynolds has listed reasons why it is preferable to have the field magnets in the guideway and the armature coils on the vehicles.


The safety of all systems has yet to be demonstrated satisfactorily, particularly that of their switching operations and during weather extremes. However, AHS is inherently less safe, as no lateral support is provided by its guideway. Braking to avoid obstructions which may appear, or when there is snow, ice or heavy rain, could cause vehicles to slide off the guideway and/or collide with other vehicles.

The simple expedient of providing fences along narrow AHS guideways may not prevent a loss of traction in extreme weather conditions, but it would prevent the disastrous consequences. Nor would it be unduly expensive to roof AHS guideways to keep them dry and clear of ice. The fencing possibility is mentioned by Joe Palen in the first posting to appear on the debate page, who went further to suggest retractable brackets that slid along, but not touch, tracks on the guideway fences. My own thoughts are that the walls could make it difficult to open doors to escape in emergency situations, and that it may be better to have plungers or folding arms near the front and rear of vehicles, which keyed into a track along the center of the guideway. One way or another, it would be a hybrid AHS/DM system.

I have briefly discussed such low-tech possibilities with Steven Shladover, who said fences, etc, had been considered, but "the challenge we encounter when we fence off an AHS lane is: how do we provide access for emergency vehicles and crews to rescue people when they have been involved in breakdowns or crashes?" The same challenge applies to the other systems being considered. (It's not difficult to imagine special emergency vehicles running along the top of the fences, something like the straddle carriers used in container depots and timber yards. They may be less appropriate in tunnels or if lanes are roofed, of course.) Nevertheless, I hope CVHAS has a further look at such low-tech possibilities to overcome the safety concerns of AHS.

The proponents of LSM say that as both propulsion and braking would be provided by electro-magnetic forces rather than friction, traction would not be affected by weather extremes, Even if a tire blew, the vehicle would experience a rough ride, but headways would be maintained. In fact the most dangerous thing would be for a vehicle to brake independently. So what happens every few thousand years when an aircraft, cow, or brick wall drops from the sky. (I once read that a cow fell from a plane and sunk a fishing boat, somewhere in the North Atlantic, I think.) I'm not sure that all such scenarios have been thought through.


I make no apology for attempting to look some distance into the future, nor for leaning towards systems which are essentially improved highways, whether or not the principal improvements are called 'guideways'. I believe the lesson of the last century is that people want cars and have preferences regarding their size. It is unlikely there can be anything better than cars, other than improved cars which are sustainable and automated, etc, whether or not they are called 'dual mode vehicles', There are limits to the extent people can be persuaded and, love them or loathe them, the masses will decide.

(Not long after the Second World War, Ford asked customers what qualities they looked for in their cars. Inevitably the answers stressed aspects such as durability, good handling and braking. So Ford concentrated on producing a good practical car and, as a consequence suffered a period about which it was said 'Ford sold safety while GM sold cars". Subsequent questionnaires asked not what customers wanted, but what they thought their neighbors wanted. 'More chrome', which was then synonymous with good appearance, was the predominant response.)

LSM systems promise so much - higher capacities, more direct control and greater safety - that it can be expected they will prevail in the long term. Of all the contributions made to the debate page, I feel my views align most closely to those of Kim Goltermann. If, as he suggests, LSM vehicles can be produced by bolting on a few magnets to otherwise fairly conventional cars, their cost should not be much increased. The greater cost would be with the guideway. If LSM increased the capacity of a guideway by more than it increased its cost, it should have the advantage over other systems.

The more immediate possibilities for DM are harder to call. There are many dimensions, including political and financial viability as well as technical feasibility. Obviously the best prospects must lie with the systems being actively developed and promoted, and it seems only RUF, MegaRail, MicroRail and AHS are at a stage where they have demonstrated, or are about to demonstrate, some of their capabilities.  Maybe a LSM system or a AHS hybrid could be early contenders if they were being actively developed, but they are not.

My assessment is that until AHS resolves its safety issues, RUF and MegaRail have the advantage. (I particularly like the flexibility of MegaRail's approach, as alternatives are considered in order to best meet requirements, rather than trying to alter requirements to fit a particular technology. (Not so easy, as the best solution for one requirement is often incompatible with the best solution for another. That's life, I suppose.) However, if AHS should resolve its safety concerns, the balance could swing the other way. By that time there should be many cars (of a type and size people prefer) with the sensing and intelligence necessary to justify the construction of AHS guideways. No need for pallets or special DM vehicles. Steven Shladover indicates this could take 10 years. I don't  know what the Japanese have in mind.

So, these are my  opinions, for what they are worth. I hope my understanding of the various possibilities that have been mentioned are not too far from the mark. I appreciate all of the effort that is being made to develop DM systems. I don't mean this to sound trite, but I wish everyone good luck - and may the best technologies win.

Brian Dudson's book To the Cities of the Second Car Age, 1995, 368 pp, paperback, is available from him for $US 20 which includes airmail postage. Contact him via e-mail if interested in obtaining a copy at: dudson@bigpond.com

Published articles by Brian Dudson are as follows:

Intelligent Cars in Cities, ITS Quarterly, Fall 1995, Vol III, No. 2.

When cars are Clean and Clever: A Forward-looking View of Sustainable and Intelligent Automobile Technologies, Transportation Quarterly, Summer 1998, Vol. 52, No. 3.

Cities, Transport and Energy, Transportation Quarterly, Summer 2000, Vol. 54, No. 3.


Last modified: April 11, 2001