Veröffentlichungen | Publications

Feasibility, Financial and Environmental Analysis of an
Advanced Maglev-Based Intermodal System

Andreas Steingröver  /  Rasmus Krevet / Robert L. Bertini


Abstract. The authors examine technical feasibility, financial aspects of a sample application and key environmental features of a newly proposed advanced maglev-based intermodal system.


The magnetically levitated (maglev) intermodal system called “Autoshuttle” is a new transportation concept which can be built along existing major transportation arteries such as freeways. At an Autoshuttle entrance station cars, trucks and buses each enter single transparent cabins. Soon after, each cabin enters the system and travels at a constant speed of 180 km/h (112 mph) to the individually desired freeway exit. The desired destination may be changed during the journey using a voice recognition module. Requesting “take next exit,” the cabin stops there after approximately 3 minutes. The transported vehicle leaves the cabin through the opened front door. The fare is paid during the journey with a credit card. The fare is cheaper than the operating costs of driving on the parallel freeway, i.e., cheaper than fuel plus wear and tear. There are quasi-continuous departures.

During the journey, convoys with very little air resistance are formed in order to achieve very low energy consumption. Entrances and exits are as frequent as on an ordinary freeway. Exits for which the user doesn’t request an exit are passed without loss of time during the journey. At the desired exit, the cabin switches out of the convoy and brakes automatically. The door-to-door average speed is very high, even for short trips.

A 3t-maglev-prototype vehicle specially suited for Autoshuttle has been built at the Technical University of Braunschweig, Germany. Transportation of an average car by Autoshuttle corresponds with an equivalent diesel consumption of 102 mpg. For a 40-ton truck the equivalent figure would be 18 mpg. Emissions during operation and raw material consumption for construction are very low. The space consumption is up to 3.6 times lower than that of a freeway. The capacity of a single Autoshuttle line corresponds to that of a 15-lane freeway.

An economic study for a 56 km (35 mile) long line along an existing freeway in Germany shows exceptionally good economic viability. Assuming a 28% mode shift from the freeway to Autoshuttle, an unsubsidized company building and operating this line would obtain high profits. A preliminary acceptance study has yielded a much higher expected value for the mode shift percentage. Profitable operation is possible on short lines already. Compared to automated vehicles driving autonomously, this intermodal system demonstrates substantially higher degrees of safety, environmental friendliness and economy.


One of the most serious transportation problems world-wide is the abundant traffic congestion in urban areas and in ecologically critical corridors. All recent traffic predictions show that motor vehicles will dominate future travel volumes. Developing countries will experience tremendously increasing traffic volumes and forecasts for industrialized countries also predict increasing road traffic. The reasons for this are the flexibility, comfort and independence and generally acceptable cost of road transportation. The road user experiences this problem mainly in the form of traffic jams and increased accident risks. Roadside residents and the environment suffer from the physical space requirements (property required to construct or expand a highway), neighborhood division, noise, energy consumption, emissions and accident risks.

New technologies such as alternative propulsion concepts for highway vehicles and telematics are very useful, but major attenuation of traffic problems cannot be expected as long as the dominance of individually operated highway vehicles is maintained. For instance, the capacity of many highways is reached with conventional traffic control, so that in these cases an automated traffic control system would only lead to a modest decrease in specific physical space requirements. Furthermore, consumer behavior still shows that safety, power and comfort are preferable features of their car and not extremely low fuel consumption.

Therefore, alternatives departing from the basic principle of individually operated highway vehicles moving on major transportation arteries have been proposed in the past. Railway trains for transporting cars and trucks theoretically allow for lower physical space requirements if the line is well used. With the operating schemes realized so far, the time-consuming and costly loading and unloading of the trains and the low station density or alternatively low average speed due to frequent stops end up leading to low traffic volumes. Additionally, the energy savings of a train system is quickly absorbed if patronage is poor or if the travel speed is considerably higher than the typical road traffic speed.

An alternative solution is the convoy concept developed by Volkswagen during the 1980s for densely used freeways (1, 2). In this system, the driver enters the slow (right hand) freeway lane and transfers control of the car to a computer by pushing a button. The car is steered to the passing (left-hand) lane and joins a platoon, so that the car forms the new front end of the platoon. Using sensors, the cars follow one other at a distance of 2 meters. During the journey, additional cars join the front of the platoon. The driver requests an exit by pushing another button, and the car leaves the platoon toward the right lane. The driver then resumes control of the car. The gap remaining in the platoon is closed automatically by the following vehicles. With this system, freeway capacity is increased, and air resistance in the platoon is diminished by 35% at 130 km/h (81 mph). Unfortunately, safety problems remain unresolved for this platoon concept. For instance if a vehicle in the front section of the platoon experiences a flat tire and loses control, the following vehicle will possibly be affected. DaimlerChrysler has developed a similar concept for truck platoons. The energy savings are lower due to the still high wheel friction which is a major part of the total movement resistance of a truck.

Another proposed system has been dual mode vehicles which have conventional rubber tires for highway operation and an additional suspension system for track guidance. With this type of system the traffic density can be increased but there is little impact on energy consumption. A disadvantage is the need for specially designed vehicles, excluding conventional vehicles from participation in the dual mode traffic stream.


Maglev Vehicles


The safety problems of the platoons formed by automated highway vehicles are avoided if vehicles are transported by maglev track guided cabins. Passengers may remain seated in their vehicles.

Figure 1 shows a maglev vehicle. The car body and the hinged front exit door are transparent, while the two laterally hinged rear entry doors and the bottom part are opaque. Solar cells are mounted on the roof and provide cooling of the cabin if necessary. The front is streamlined and the rear part of the cabin extends over the rear doors. During the journey in a convoy the following cabin closes up directly to the end of the preceding cabin. Since the cabins fit together in a modular fashion, a streamlined, nearly smooth transition between the cabins with constant cross-section is achieved.

Figure 2 shows that the cabin sides pivot to form auxiliary doors so that the passengers may leave the highway vehicle or the cabin in extraordinary circumstances. Remote controlled ventilation windows are also provided. There are cabins with small cross-section for passenger cars–2.20 m (7.2 ft) internal width and 1.70 m (5.6 ft) internal height–and cabins with large cross-section for trucks and buses–3.30 m (10.8 ft) internal width and 4.30 m (14.1 ft) internal height. Both types are provided in different lengths–from 3.60 m (11.8 ft) to 5.60 m (18.4 ft) internal length for cars and from 6 m (19.7 ft) to 19 m (62.3 ft) internal length for trucks and buses. All types ride on the same track and form convoys from cabins with identical cross-sections. The typical operating speed is 180 km/h (112 mph) for all convoys.


The uniform speed yields an optimal line capacity. This speed is below what is technically possible but has been found to be sufficient to make Autoshuttle transportation clearly faster than conventional road traffic. At this speed, energy consumption is very low, noise is almost negligible (see below) and relatively sharp curvature is acceptable (minimum radius 1250 m). In extremely congested areas a speed reduction would be possible in order to combine Autoshuttle with very sharp curves of an existing highway right-of-way. A gradient of 10% yielding short ramps can be traveled at a constant 180 km/h (112 mph).

Inside the cabin is a flat movable communication module mounted on the driver’s side. The module automatically moves towards the opened driver’s window. The driver uses the communication module to enter the desired exit station by voice recognition or keyboard and pays by a credit card. Alternatively, a mobile phone service can be used for this purpose. The type of highway vehicle is determined at the entrance station by a number plate identification system using a vehicle registration database. The fare is calculated based on vehicle type including type of engine and the corresponding operating cost. The fare is set at a point 15% below the average cost that would result from driving the highway vehicle under its own power, i.e., the cost of fuel, oil, wear and tear and mileage-dependent depreciation determined for each vehicle type. The highway vehicle’s dimensions are determined by light beam detectors, so that a suitable cabin is ordered. Furthermore, a fast exit button for exiting at the next station, an emergency call phone, a power supply for the highway vehicle’s equipment and a cabin ventilation and window remote control are provided to the driver.



Figure 3 shows a station plan. Stations are located at approximately 5 km (3 mile) spacing, on the order of freeway interchange spacing. Via a passive switch (location 1 on Figure 3) an exiting cabin, 2, leaves the convoy, 3, (the operation of the passive switch will be described below). The vehicle brakes on a 1 km (0.6 mile) deceleration track, 4, turns to the right at point 5 and stops in an exit bay, 6, where the highway vehicle leaves the cabin through the front door under its own power. Thereafter the cabin moves backward toward an entrance bay, 7, where another highway vehicle enters. As soon as a convoy, 3, has reached a reference position on the main track, the freshly loaded cabin accelerates, switches on to the main track, 8, via a passive switch, 9, and is swiftly caught by the convoy, 3, upon reaching the operating speed.

The cabins that do not wish to exit pass the station at full speed. Average speed is therefore nearly 180 km/h (112 mph). The car convoys follow one another at 2-minute headways, while truck and bus convoys have 6-minutes headways. The frequency would decrease during the nighttime hours. Physical coupling of the cabins is in principle not necessary, however simple engaging couplers which uncouple using lateral motion are provided. The convoy, 3, need not be stretched when a cabin leaves the convoy, 3, at the passive switch. At interchanges cabins can change Autoshuttle lines automatically.

Supporting and Guidance System and Passive Switch


Figure 4 shows the experimental mock up of a maglev vehicle and Figure 5 shows a detail of the right hand side magnetic levitation and guidance system from the front end with two L-shaped rails on each side of the cabin. The levitation bogies of the cabins enter between the two rails on each side and engage from beneath the rails. Symmetric magnetic circuits with minimized energy consumption are formed by a permanent magnet which is controlled by an excitation coil and the rails. The configuration of the levitation system enables the levitation function even when one rail per side is omitted.


This is the case on some parts of the passive switch as shown in Figure 6. In addition, lateral movement control magnets are provided, which are activated for short periods when entering a passive switch. For example cabins turning to the left activate the control of the additional lateral movement control magnets. The cabin travels contact-free by its on-board magnet along the right-hand branch of the passive switch.

As an additional mechanical safety device, vertical guidance rails are mounted at the switch in the center of both the straight and deviating branches. Under the cabin at the front end a guidance pin is opposing; this pin can move laterally. The cabin approaching a diversion point determines the intended direction before the braking distance of the switch is reached by activating the additional lateral motion magnet as described above and by moving the guidance pin in the desired direction. The pin is latched at the end position. An emergency brake is applied on failure. The guidance pin travels contact-free laterally along the guidance sheets. Erroneous guidance is not possible even in the case of magnet failure due to the presence of this engaging mechanical safety device. Therefore, the safety standard of this passive switch is at least as high as with conventional switches (dimensions of the passive switch are available at


Propulsion System and Rendezvous Maneuver

Autoshuttle has a long-stator-linear-synchronous-drive with an iron-free stator (3) winding placed beneath the rails on each side of the track. In track sections, where cabins move with very small spacing from one another at different speeds, motor sections reach short lengths down to 2.70 m (8.9 ft). Each of the short motor sections is fed by a power inverter with corresponding pole position sensors and motor current control. The motor has a simple configuration and reaches high efficiencies due to the low power demand of the convoys at constant speed and due to the short motor sections during the accelerated motion. Power demand reaches 150 kW per meter (45 kW per ft) for accelerating a cabin containing a heavy truck.  During travel at constant speed, the power demand goes down to approximately 4 kW per meter (1.2 kW per ft) for a heavy truck cabin and 2.5 kW per meter (0.75 kW per ft) for a passenger car cabin.

The rendezvous-maneuver is enabled with the individual control of the short motor sections. The control principle becomes quite simple if predetermined curves for the movements of the approaching vehicles are used. Small deviations are corrected by the motor control. Only larger disturbances or defective motor sections require an adaptation of the predetermined curve.

Control and Safety System

A control center maintains control of the operations. Communication between the cabins and the control center takes place by radio or high frequency leaking cable in the track bed. The control center receives the following information from the cabins:

The cabins receive the following information from the control center: The control center processes the information received from the vehicles and provides corresponding direction commands for the cabins. The track contains Hall-sensors detecting the presence of cabins. If the sensors detect that a vehicle remains behind its intended position, all following cabins, which could come into a conflicting position with this cabin, will be braked after a tolerance interval. The control center calculates track occupancy after the passage of a passive switch according to the direction indication of the cabins issued earlier. Indications of desired exit stations are used for the coordination of the empty runs required for dispatching the necessary number of cabins to each station. In addition, a daytime and calendar-dependent forecasting program is used for this purpose. In order to save energy, empty cabins are dispatched with loaded cabins whenever possible.


Energy Consumption

The Autoshuttle’s energy consumption includes cabin consumption due to:

Air resistance has been calculated by two methods (detailed calculations of the air resistance and the resulting energy consumption are described at

1. Application of the air resistance formula for rail vehicles of the Deutsche Versuchsanstalt for Luft- and Raumfahrt (German Research Company for Air- and Spacecrafts).

2. Numerical analysis using aerodynamic similarity to the maglev-vehicle Transrapid TR08.

Both methods yield an aerodynamic resistance coefficient cw = 0.69 for a 177 m (580 foot) long convoy with 38 cabins for cars (this assumes a 5.8 m2 (62.4 ft2) cross section, average cabin length of 4.6 m (15.1 ft), and that an empty tail car with a streamlined form could be added at the end of the convoy; similarity calculations based on (4, 5)). The value diminishes for shorter convoys and reaches cw = 0.28 for a single cabin. Eddy current losses in the rails strongly depend on the choice of material and the distances between the cabin-borne supporting and guiding elements of each cabin during the journey in a convoy. It is assumed that a convoy 177m (580 feet) in length bears a propagation resistance caused by eddy currents of 10% of the total propagation resistance. This value is doubled for cabins travelling singly.

On-board energy demand is caused by the highway vehicle’s equipment, the air gap control of the levitation system, the communication module and the cabin window control. The highway vehicle has a power demand for heating or ventilation and further equipment of approximately 1.5 kW.  The air gap control requires 0.2 kW/t. With the vehicle’s empty weight of 3 t and a load of 2 t the gap control’s demand is 1 kW. Further on-board equipment yields an averaged 0.2 kW. Averaged on-board equipment consumption therefore adds up to 2.7 kW.
A typical realistic journey with the following parameters will be examined (detailed energy consumption calculations are available at

Empty runs are included in order to dispatch the cabins (calculation of energy consumption considers empty runs).  These empty runs have the following parameters: Every 5km (3 miles) there is a station, each having a power demand for illumination, cabin door actuation, shunting movements and optical recognition systems of 20 kW.

Efficiency from entrance to the sub-station and the motor air gap is dependent on the power demand of the motor section and the coverage ratio of the vehicle length to the length of the activated motor section. For the typical journeys as described above efficiency varies between short term 70% during braking and 91% during the journey of the convoy at a constant 180 km/h (112 mph) on level terrain. Average efficiency from the power plant to the air gap is assumed to be 32%. This yields primary energy consumption of 24 kWh per average car per 100 km (62 miles). This corresponds to a comparison value (primary energy consumption contains 8% for refineries, infrastructure and transport) of 102 mpg of diesel fuel. Analogous considerations yield, for example, 18 mpg for an 18 m (59 foot) truck. Assuming that electric power is furnished by coal, gas or fuel oil power plants and long distance heat supply is realized, the primary energy consumption is further reduced by 40%.

Resource Consumption

The resources consumed considering the construction and operation of Autoshuttle are estimated and compared to the results of ordinary highway traffic (see The conclusion is that the Autoshuttle system consumes dramatically fewer resources than a highway traffic system.


Table 1 shows the emissions that result for passenger transportation. The results are compared to ordinary car traffic and the German Railways high-speed train ICE.

System  l/100Pkm CO2 CO HC NOX SO2
ICE train
Autoshuttle car
Autoshuttle bus

Patronage is assumed to be 1.7 passengers per car for the car and for the Autoshuttle. For the ICE, data from German Railways (6) were used. The units for energy consumption in column 2 are liters of diesel fuel per 100 passenger-kilometers. The units for emissions are g/100 passenger-kilometers. The calculations are based on German electric energy production modal split, however with the caveat that no radioactive emissions occur, since the emissions values of the Autoshuttle and the ICE were augmented assuming that no energy was produced by nuclear power plants. To conclude, Autoshuttle’s emissions are much lower than those for cars and high speed train ICE.


Related to the measurements of the TR 07 maglev vehicle noise emissions of a convoy at 180 km/h (112 mph) of less than 74 dB at 25 m (82 feet) distance can be expected. With typical convoy frequencies this yields a very low average noise level, making noise reduction measures generally unnecessary.


At capacity, the main line is fully engaged by convoys except for gaps required for entering cabins and safety tolerance intervals. Passenger car convoys operate at 2-minute headways and truck and bus convoys at 6-minute headways. The result is a capacity of 15,000 transported highway vehicles per hour per direction or 30,000 highway vehicles per hour on a double lane (details of the calculation of maximum traffic capacity are presented at This corresponds with the equivalent capacity of 15 freeway lanes. The overall physical space requirements, i.e., for track, stations and storage yards is then 3.6 times lower than that required for equivalent throughput on a highway (details of the space consumption calculation are presented at  To handle the traffic of one six-lane freeway, the physical space requirements of the Autoshuttle are half of that of the freeway.


Assuming an initial scenario of a congested six-lane freeway, which is being considered for expansion to eight lanes. Here is a case where Autoshuttle could be built instead of the widening project. Figure 7 shows a combined four-lane freeway and Autoshuttle station. If the Autoshuttle indeed generated substantial demand, Autoshuttle’s main tracks could be built on the freeway right-of-way, reducing the freeway to four lanes, which would be sufficient due to the lower remaining traffic volumes. The combined structure would have the same space consumption as a conventional eight-lane freeway. The vehicle-carrying capacity would equal that of a ten-lane freeway and could easily be increased. Station location would be flexible because highway vehicles could travel short distances to the next station.

This yields the interesting prospect of designing an Autoshuttle in the median of a freeway without needing to widen the cross-section of the combined facility at extremely space-critical sections. The loading and unloading capacity of a bay has been estimated based on practical tests of the average time to enter a garage with similar dimensions as an Autoshuttle cabin. It was estimated that 109 cars or 63 trucks and buses could be loaded per bay-hour. Thus the average station is quite small with typically six loading bays and six unloading bays per direction. This relates to a six-lane freeway with 10% of the traffic flow using the entrance. A large station, e.g. close to a stadium, would typically be dimensioned with 18 bays per direction and per type, having a total unloading capacity of 4000 cars per hour. The same value applies to the loading capacity. Principally, cabins could be routed to adjacent stations in case of excessive demands. The affected cars would then drive to the desired exit.


Potential user acceptance has been assessed by a preliminary nearly-representative survey among 135 people (survey details are provided at  The question asked was: "Would you use Autoshuttle instead of an ordinary freeway?” Important parameters are:

Respondents have answered “yes” 95 % of the time. Interesting aspects relating to sensitivity are:


An economic study has been conducted for the sample line between Breitscheider Kreuz and Köln Airport (Germany). The Length is 56 km (35 miles). The objective of the calculation is to determine the minimum percentage of highway vehicles switching from the parallel freeway over to Autoshuttle in order to make possible a subsidy-free profit for the building and operating company.

According to the lowest prediction,  (7, 8) an average of 124,000 highway vehicles will travel on this freeway per day in the year 2010, the assumed inauguration date of Autoshuttle. The fare for cars, trucks and buses is set at a point 15% lower than the cost of driving on the parallel freeway for each vehicle type. This results in average fares of 10 cents/km for cars and 23 cents/km for trucks and buses (detailed calculations for what a car owner would save by driving one kilometer less are provided at at 2000 German price level. Expenses for the construction of the line are estimated referring to the cost estimates prepared by Thyssen, Siemens and AEG for the Transrapid-maglev-line between Hamburg and Berlin as well as cost tables for German Railways construction (a detailed description of the financial model is presented at  Autoshuttle is conceived of as being financed entirely privately, without public subsidy. Summarized results are shown in Table 2.

Costs (first year)  
Financing Cost (25 year term) $79 Million
Operating Costs $34 Million
Total Cost $113 Million
Revenue (first year)  
Based on daily Autoshuttle volume 35,037 veh. $113 Million

These figures represents a mode shift of 28% from the freeway to Autoshuttle. Considering the promising result of the preliminary survey, this mode shift could be exceeded. The cost coverage ratio rises in the following operational years and reaches 260% in the 26th year of service. Autoshuttle can be operated on lower volume routes as well.

This analysis has been conducted based on conditions in Germany.  Adaptation of this economical study to the situation encountered in U.S. yields the following main differences:

All factors combined yield a minimum changeover rate of the same order of magnitude. Autoshuttle therefore shows excellent financial aspects for a U.S. application as well The total mileage of roadways worldwide where Autoshuttle could be built and operated without subsidies and with profit exceeds 60,000 miles.


The proposed new transportation concept Autoshuttle is capable of mitigating the problems of abundant road traffic. Autoshuttle permits use of conventional highway vehicles and is:

In conclusion, the new combination of features for this intermodal track guided transportation system include: AUTHORS

Andreas Steingröver:   DaimlerChrysler Rail Systems (Signal) GmbH, Email:
Rasmus Krevet:   DaimlerChrysler Rail Systems (Signal) GmbH, Email:
Robert L. Bertini:   Portland State University, Department of Civil Engineering, Email:


1. Fiala, E.: Automobiltechnische Zeitschrift ATZ 90 (1988), 429
2. Dreyer, W. et al.: VDI Berichte 699 (1988), 311
3. Steingröver, A.: Stützung and Antrieb eines permanenterregten Nahverkehrsfahrzeugs, PhD thesis, Braunschweig, 1996
4. Fürst, R. in: Proceedings of the 14th Int. Conf. on MAGLEV, Bremen, Nov. 26-29, 1995, VDE-Verlag Berlin (1995), 113
5. Miller, L., Löser, F. in: Der spurgeführte Verkehr der Zukunft: Tagung Hamburg, 18./19. Juni 1998/VDI-Gesellschaft Fahrzeug-und Verkehrstechnik, Düsseldorf: VDI Bericht 1392, (1998), 136
6. Jänsch, E.: Elektrische Bahnen eb 93 (1995) 1/2, 25
7. Traffic Prognosis for Germany, issued by Prognos Institute, Basel, Switzerland, 1993
8. German Federal Traffic Lines Plan BVWP 1992
9. Traffic Counts in Germany 1996, File dtvj96.txt, issued by the German Federal Ministry of Transportation, 1997