The Train Wire: A Discussion of the Science of Train Dispatching / (Second Edition)

by John Alexander Anderson

The closing paragraph of the rule provides for the contingency of the failure of telegraphic communication at a critical moment in the transmission.

An order may have been fully received by an operator, but, if the telegraph fails before he can repeat it back and be informed by the Dispatcher that it is "O K," it would not be safe to use it. Neither is it proper that it should have any effect whatever until the Dispatcher is assured, by the acknowledgment of the "O K," that it has been received. When an order has been transmitted and is altogether in the hands of the operator, there is the chance that he may have written down some important word incorrectly. Hence the requirement that he repeat it back. This, if carefully performed, assures the Dispatcher of the verbal accuracy of the message as the operator has it, and the Dispatcher admits this by the response "O K." He must now act, with reference to this train, as if it were held at the point at which it is addressed. But he cannot assume this until he is assured that "O K" has been received. This is by the required acknowledgment.

If communication absolutely fails before the completion of this process, all that he has done goes for nothing unless communication is quickly restored. It is of the utmost importance that the Dispatcher know what will or will not be done by a train to which an order has been addressed, as this knowledge guides him in giving other orders. It would not be proper, even, to assume that a train would be held by the presence of an order addressed to it unless the accuracy of the order is assured, for an error may have occurred in receiving the address and the wrong train number may have been noted. Nor will it do for a train to proceed regardless of an order addressed to it when the whole process of transmission cannot be completed, unless the rule authorizing it is made to specify the precise point in the process of transmission when this may be permitted. It is also of equal importance that, in the absence of telegraphic communication with a train, the Dispatcher can depend upon the fact that it will act in accordance with the rules, notwithstanding a partial transmission of an order intended to control its movements. Briefly, he must know whether the train retains the right to proceed or not, and under what conditions, or he cannot intelligently direct other trains with reference to it. The question how long a train should wait for communication to be restored must depend upon so many circumstances that no rule can be given. The "break" may be but momentary or it may last for hours. The train may have just time to get to a regular meeting-place, at which, if reached in time, it may have to lie for belated trains. Rules must fail here to indicate what is best to be done, and often the best judgment is no guide. Whatever is determined on may involve delay. It should never involve danger.

There is a plan in use on several prominent roads by which it is claimed that the objectionable feature in Rule 510, represented by the phrase "whenever practicable," may be eliminated. Under this plan there is added an "advance" order, issued to the superior train, directing it to stop "for orders" at a point where it is intended to deposit for it the duplicate of a meeting or other order on which an inferior train is to be permitted to proceed from some other point before the order is received by the superior train. By this plan the superior train is "held" before the inferior is allowed to act on the order, and thus far the risk is avoided of the superior being improperly allowed to pass the point where the duplicate order is to be placed for it. It is claimed that a considerable experience has demonstrated that this plan is feasible and secures the object in view, and that with it the rule of always first securing the superior train may be made absolute. Experience is one of the best of teachers, and few theories can be taken as proved without it, but even imperfect methods may produce good results under careful management, so that experience alone is not sufficient for determining the merits of a system.

The purpose of the plan in question, to "hold" the superior train before giving orders against it is good, and what all wish to accomplish. This idea gave rise to the "hold" order of the older methods of train dispatching and it has been suggested that under the advance-order plan there is danger of a relapse from strict adherence to the duplicate method. Careful supervision may prevent this.

If the advance order is invariably given, operators may get to depending on it rather than on their own care for stopping trains at points where duplicates are deposited. This is a point to be carefully considered and on which the railroad fraternity will be by no means agreed. Two things are depended on. If one fails we have the other. Many hold that this is better than to rely on one alone. Many, again, maintain that, where the responsibility is thus divided, each party may depend on the other and both fail, while, if there is but one, his sense of responsibility is quickened and the result is better. In view of the difference of opinion on this point it may be said that if this be the only point in the consideration of the advance order it may be given a trial.

If it is to be tried, then we must see that there are no exceptions to its use. The Dispatcher must always anticipate possible contingencies long enough ahead to be able to designate in advance the points where trains are to stop for orders, and he must do this before the necessity arises of allowing the inferior train to proceed on orders which the superior trains are subsequently to receive. If he cannot thus anticipate he must still give the order to stop for orders and send it to the point to which the meeting-order is sent, both to be delivered to the superior at the same time; and in that case he must depend upon the signal at that point for stopping the train, as in the Standard rules, or always keep the inferior train from acting on the order until the orders for the other train are delivered.

Again, a train for which it is thought meeting-orders may have to be given must make a stop in order to get the advance order, and again another at the point named in it, perhaps only that it may receive an order annulling the first, if meeting-orders are found not to be needed. Frequently a duplicate order may be placed for a train and annulled before its arrival if the occasion for it has passed, but the advantage of this is lost if the advance order is used.

There are many roads on which the circumstances would not admit of thus always seeing far enough in advance the things to be done, and very many on which the business would not admit of the stops necessary, and the occurrence of a single exception would vitiate the whole and make it necessary to fall back on the provision "whenever practicable."

It is not easy to see how the rule could be invariably applied at junction points at which trains of superior right are to arrive from other roads or divisions, and circumstances are so various that it is difficult to determine just where such a plan could or could not be satisfactorily applied. Some say they have succeeded with it. Others point out quite conclusively that the circumstances with them are such that it would be impracticable. Where it can be applied and used without exception and the question of divided responsibility can be satisfactorily disposed of, it is, to say the least, an experiment in the right direction, but it is to be very much feared that this plan does not yet supply the universal remedy for the difficulty involved in the phrase "whenever practicable." The multiplication of messages on a busy wire will occur to all as a serious objection, but scarcely as one that should weigh against positive considerations of safety.