A Reply to Kim Goltermanns Comments on an Individual Rapid Transit System
William G. Turnbull
Mr. Goltermann has indicated that I believe that because a pallet does not leave the guideway, the condition of the pallet is better known. Not quite. The condition of the pallet is better known because the pallet is completely under the control of a single agency; thus it can be taken off-line for scheduled, periodic inspections and preventive maintenance. He claims that this is not necessary since a computer can monitor all the critical aspects of any vehicle, and that simply by interrogating the computer, one has all the information needed.
I remain unconvinced. Moreover, I find my self in good company. For instance, a modern airliner is probably the most computerized piece of machinery extant today. Yet the FAA (US Federal Aviation Authority) still requires periodic withdrawal of all aircraft from flight operations for complete inspection and preventive maintenance. You may be sure that if it could be shown the computers could do the job while the aircraft is in service, the airlines would insist on it. Taking airplanes out of service is very expensive. But even under the strict regulation of the FAA, incomplete or falsified inspections have lead to failures, which resulted in fatal crashes. At least initially, inspections must be under the direct control of the responsible agency. An early fatal accident could prove disastrous to the dualmode idea, no matter of what stripe.
Computers can not examine for micro-cracks in critical structures such as bearing races, axles and the like. It can not easily look for cracks and suspicious wear patterns in tires. Nor can it identify abnormal wear and other indicators of potential problems in other areas. In short, until it can be shown conclusively that we do not need an experienced pair of eyeballs in the loop, prudence demands it. Merely the assertion that it is no longer required does not make it so. Reality requires verifiable facts, not assertions.
Moreover, I find the argument that because the wear and tear on vehicles may be harsher on the street than on the guideway, failure thus is guaranteed to occur on the street simply lacks credibility.
Mr. Goltermann assumes that "dualmode cars will not deviate much conceptually from the cars we already buy and drive." Thus, presumably, the overwhelming majority will forgo the advantages of electric propulsion. In any event, he anticipates they will incur little or no additional cost.
Whereas, a pallet will be an over-engineered behemoth requiring in addition to the obvious brake and propulsion systems such extraneous systems such as guideway shifting mechanism, sensors, back-up power source, coupling devices, communication systems, computers, maybe even AI. These things are apparently not required in true dualmode. Perhaps he can explain this, I cant.
Mr. Goltermann makes much of the fact that true dualmode vehicles will be in greater abundance than a pallet would. Thus he suggests that a pallet might cost as much as ten times that of dualmode vehicle. Economics, like engineering, is a business of verifiable numbers. Once the quantity becomes sufficiently large that the amortized engineering is a small part of the cost, quantity tends to irrelevancy. If one can amortize the engineering over a 100 thousand vehicles, the fact that one million are built does not reduce the cost by a factor of ten. It is more likely that the effect of an increased bureaucracy will take over and tend to drive the cost back up.
He also stresses that fact that storage of not-in-use pallets will be required, and that the cost of a storage facility will be in the same order as a main line guideway. On this latter point, he is wrong. These can be stored in nothing more glamorous as a parking lot, with simple buried electrical cables for guidance. The location can be specifically selected to minimize real estate and other costs. That there will be a requirement for additional pallets beyond that required accommodate peak revenue traffic is undoubtedly true. Whether it reaches a number as large as 50 percent is currently being studied. I rather doubt it. But unlike Mr. Goltermann, I prefer to await the results of a rational study, rather than simply throw out numbers.
That this represents an additional financial burden on the specific initial cost to the operating agency is also true. However, whether it represents higher cost to those who will ultimately pay for all this (i.e., the commuter) is not nearly so clear. In this connection, it interesting to note that one suggestion offered to solve the "chicken and egg" problem is to offer a true dualmode vehicle approximately 2 years before opening the guideway. The idea of a new car every two years is no longer a reality in the US, so it is not clear how effective this might be. But if it were, it would also mean that the additional equipment required for dualmode operation would be re-purchased every two years. As it is, it is quite likely that new cars (along with the dualmode requirements) will be purchased far more frequently than pallets would need replacing. Thus, not only would we be forcing the commuters to purchase this equipment many times over what would be required with pallets, but also we are requiring them to do it more often. Yes, the additional equipment might be transferred to the next vehicle, but this would defeat the very argument of mass production that Mr. Goltermann advances.
It is also interesting to note that Mr. Goltermann concedes that "temporary" pallets may be required to counter the chicken and egg problem. However, unlike "real" pallets, these temporary ones could be compatible with a true dualmode system, whereas real one could not. The reasoning for this escapes me, but it really doesnt matter. The thrust of my argument about pallets is that they are required initially
I am also somewhat dubious about the relying on automobile companies to set standards that are optimal for the individual user. They do not have a particularly good record with either safety or environmental issues. Moreover, I am reluctant to proceed to operational standards of any sort, absent any experience with this type of system. Note that I refer to operational standards, not safety standards. I believe a palleted system provides the greatest flexibility for development and universal use, and thus will lead more surely to the development of an optimum system.
Second generation systems will no doubt differ considerably from their predecessors. At that time, with real operational data in hand, operational standards can be considered, including, perhaps, true dualmode - but not yet.
On the issue of control, and in particular coupled packets, capacity was not the driving issue (although it helps marginally). The principal reasons were safety and ease of control. If separate vehicles are mechanically coupled, there is no opportunity for relative kinetic energy to build up between the vehicles. This can be of great significance in an emergency stop. Moreover, by concentrating space in the headway between packets, it is much more likely that subsequent packets can be halted without incident, again each as a single entity.
Moreover, if the motor or brakes fail in a single vehicle, the remaining vehicles in the packet can take up the load and bring the entire packet off the system at the next exit.
I see no particular difficulty in coupling or uncoupling at speed as long as the relative velocity is kept small. I see no greater difficulty in approaching a vehicle to couple than in maintaining a specific, small distance behind. Admittedly, when these operations are ongoing there is some increased risk, but no more so than what would be experienced continuously when running in close proximity. It is better to be exposed to risk a small fraction of the time than throughout the entire journey.
In developing the packet idea, a great deal of emphasis was given to minimizing the amount of data to be stored and/or transmitted about. With a packet, it is required only to know the number in the packet to know a location for each member. Mr. Goltermann is somewhat sketchy about the details of his system, about what criteria are used to effect switching and the like. However on the face of it, it appears that maintaining individual vacancies (individual packets), would suggest that something like 20 times the data must be maintained. When considering expanding the system throughout an entire region, this is of critical importance.
He quotes a probability of 98.85% of finding at least one vacancy in twenty. That, as he points out, is the probability for a single vacancy. At busy intersection, within a span of 20 vehicles there is frequently the requirement for several empty slots, not just one. He states that "The figures [he has] been using indicate 3 5 alternate slots will be plenty . . ." He provides no information concerning the assumptions made, how he arrives at that conclusion, or the source of his information. If the term multiple slot clear path is to be taken literally, he should provide some indication of the algorithm he intends to use to establish and maintain a clear path, together with an estimate of the total computer power required to implement it. It would also be useful if he provided some guidance on the consequence of failure.
If not a clear path, he should provide information on the frequency of conflicts, and how he would manage them - or does he assume that they would never occur. Unless or until he provides some greater detail, I find it difficult to take his arguments seriously. He does suggest that some type of quota system will ameliorate some of these problems (and I am grateful for the attribution); but again he provides no hint how it might be incorporated.
I have provided both calculations and detail at: http://www.poly-soft.com/irts It describes the approach I used to address the critical issues of safety and capacity for a dualmode system. I invite Mr. Goltermann to do the same. Then all can judge for themselves, which is the superior system. Or quite possibly, it will inspire someone else to devise an even better system.
Last modified: September 15, 2002