SOMETHING OLD, SOMETHING NEW
Page 73
Page 74
If you've noticed an error in this article please click here to report it so we can fix it.
THE main purpose of the paper pre
sented by Mr. G. G. Harding, general manager of Wallasey Corporation Transport, was to describe the method of operational research carried out in his undertaking. Mr. Harding introduced his paper by covering the now all-familiar ground of increasing car usage, traffics congestion and drop in the number of passengers carried by public transport. It was a sobering thought that had a cheap 1965 minicar been perfected in 1890 none of us might ever have seen a bus.
If public transport was to form a large part of the future means by which people moved it would be necessary, at some stage, to calculate how public funds should he divided between roads, car parks, buildings, and transport systems and to what extent people who use these services should directly pay for them in order to ensure that the greatest good was done for the greatest number. How was this to be done, and how did we provide a measure of what was "the greatest good"? Again who, today, was quali
fied to specify what constituted "social benefit "?
Nearly every normal remedy used by transport management in the circumstances described had been tried long before he had arrived in Wallasey, said Mr. Harding. The transport committee had reached the conclusion that re-routeing of the Borough's services was necessary. bearing in mind the need to integrate them with the Corporation's ferry service from Liverpool, and it seemed clear that any new approach would need to include the investigation of this matter as a primary objective.
One method considered was the use of market research techniques and it was thought that this would be good and cheap if carried out on a "do it yourself" basis. There were, in fact, many shortcomings in this approach, some of which were well known. Even if one was fortunate enough to be able to afford a market research investigation it had to be remembered that even having obtained accurate information, which was in itself difficult enough, it was very easy compared with using that information. It was as a result of one of the many letters of inquiry sent out that the undertaking
heard of the term " operational research" for the first time.
Mr. Harding then went on to describe at length the function of operational research (OR). A preliminary investigation of Wallasey's particular problem was undertaken by lecturers from the OR departments of Birmingham and Hull Universities. These lecturers were asked to determine whether OR could successfully be applied to the job of tailoring services to present-day requirements. As far as could be ascertained OR had not been applied to such a problem anywhere in the world before, so that new ground was being broken from the outset. The lecturers' report, prepared after a short intensive investigation, stated that in their view an OR solution would be possible and recommended a Pilot Operational Research Project. The transport committee and Council accepted this recommendation and the Pilot Project was carried out by Birmingham University within a period of four months.
Over 10 Years A period of 10 years was taken as a basis of comparison to compare conditions of today with those formerly applying. It had often been said to him, went on Mr. Harding. that if the period had been different the position would also have been different—which, of course. would have been true. However 10 years had been chosen somewhat arbitrarily as they covered most of the period during which the undertaking's traffic had suffered a post-war decline. In any event. it was felt to be of little consequence to either management or the passenger to know that bus rides are today relatively cheaper than they were when Adam was a lad; even if it were true no one who used buses could ever be persuaded to believe it!
It was well known that more often than not the car owner quite wrongly considered the cost of a journey by car as being only the cost of the petrol consumed. Today, while he could ride three miles for 8d. on a bus—once he had walked to it and waited for it and possibly fought for it, too—if he had a small car he could go at least four miles for 8d. worth of petrol without walking, waitor fighting. The fact that the four miles by car had certainly cost him 2s. did not impress him one little bit.
During the Birmingham University Pilot Project in Wallasey a route-finding computer programme was written. designed to be fed with information regarding the journeys actually taken by passengers so that the routes produced would be in accordance with actual passenger demand. It was necessary to establish throughout weekdays and week-ends where passengers boarded buses, where they alighted, by which route they travelled and at what time. This the undertaking did not know.
The method chosen to obtain this basic information was to use a number of ticket machines containing audit rolls so c46 that the required data could be read off in respect of all passenger journeys for which tickets had been issued. This information was analyzed and built up into a picture of the point-to-point demand pattern or matrix for each of the several periods of weekdays and weekends.
It was recommended in the report on the Pilot Project that a full scale Operational Research Project should be carried out by one of a number of commercial organizations capable of doing so to continue the work which had been done to the, point where practical results would be achieved. The Council accepted the recommendations of the report and a commercial organization started the further work in Wallasey in October, 1964.
For this full-scale project, audit-type ticket.machines were used in substantially larger numbers than before because this was considered the cheapest and quickest way of obtaining the necessary data, consistent with the degree of accuracy required. This information was supplemented by (a) employing personnel on buses to ask passengers if they were going to change to another bus to complete their journey and. if so. where and what was their final destination; (b) employing personnel to observe changes at known transfer points and their destinations after alighting at various points-, (c) the use of house-to-house questionnaires to a limited extent where this was unavoid-. able.
After considerable discussion the criterion chosen for judging one set of services against another was that of the total travelling time of all the people making journeys by bus for a given total cost. The travelling time of one passenger was the time taken by the passenger walking to a bus stop. waiting for a bus. riding on the bus to the alighting point and walking from there to the ultimate destination. The walking time was calculated at different speeds in hilly areas and flat areas; where a change of bus was required to complete a journey. additional factors were taken into account, such as the length of wait for the next bus and the annoyance factor of having to change buses.
The total travelling time was the summation of all these factors for all the people who travelled or would travel on a given set of services. If the costs and incomes resulting from operating different levels of service which required varying numbers of buses were plotted on a graph one could see for a given fare level the number of buses which could be operated at the break-even point.
There was, of course, nothing new in this and it was with these factors that people in transport had always juggled in order to achieve a balance. What was new was the basis upon which, and method by which, the juggling was done. The new method involved the use of what was called a "mathematical model ". This mathematical model included, amongst all the other variable factors, a number of new, possible and practical routes found by the first corn puter programme. These had been included as well as the existing route pattern in order to comply with one of the prime objectives of the project which, as said earlier, was to investigate the possibility and desirability of changing the existing pattern in Wallasey.
The results for.some periods of the day had just been produced at the time of writing the paper. and one result was of particular interest as an example of a difference which comes from the approach described as compared with the human approach—the computer had sdapped at peak periods a very substantial route which was currently one of the undertaking's " best" in terms of revenue per mile!
One's initial reaction to that was horror, as it seemed that it could not possibly be right. Perhaps some vital information had been omitted or an error had occurred. In fact the reason was neither of these. In that case the machine had selected certain existing routes and certain .computer-produced routes. It had rejected the whole of some and parts of other existing routes on which receipts were low, as well as some on which they were high, because it could be shown that the new complete combination was "better" than the old. No new route directly replaced the high receipts route that it had scrapped, but the demand on all parts of it were catered for by the new combination.
The Wrong Approach?
Mr. Harding was sure that he was not alone in having wondered in the past whether time devoted to trying to make uneconomic routes less uneconomic would not have been more profitably spent in trying to make profitable routes more profitable, and the above result seemed to indicate that something of the sort was probably true.
Having collected and analyzed the large amount of data and run the various programmes; the committee and Council having decided a financial policy and. therefore, determined the set of services and frequencies, and taken other associated decisions; the department having subsequently completed the necessary licensing work, negotiations. publicity and public explanation; then one could, if the result of the project had been correctly " sold ". introduce a bus service which would do the greatest good for the greatest number of passengers—under the conditions prevailing at the time that the basic information was collected.
But, of course, nothing remained static. said Mr. Harding. Passengers' habits continued to change. costs continued to increase, the level of traffic congestion in general would not remain as it was, and so on. In order to ensure that the undertaking continued to match its services accurately to the changing passenger demand, or, alternatively, so that it could know the cost of not doing so. it was necessary to re-examine actual passenger demand from time to time. This would be done by re-sampling at appropriate intervals—say, every year. using the same method.