Savings by simulation
Page 22
If you've noticed an error in this article please click here to report it so we can fix it.
The potential of transport consultancy
• Examples of the savings which operational research can offer in road transport were given by Mr. K. P. H. Fielding, head of British Road Services Ltd. consultancy service, when he spoke at a well attended meeting of the Institute of Transport's Berks, Bucks and Oxon section in Reading on Monday. Demand for optimum transport solutions, he said, called for better answers than the "about right" which could be achieved by hunch and experience. The "just right" answer could usually be obtained by operational research, and particularly by using simulation methods.
He gave the recent example of a BRSL study for a high-intensity transport job involving maximum-capacity artic tippers and the possible future use of new max-cap vehicles in the 38/40-ton class. The "about right" answer was that 12 vehicles costing about £8,000 each could do the job. The consultancy service was then called in, and produced a scheme which would require only eight vehicles and two spare trailers, so saving about £20,000 in capital cost.
This was achieved for the expenditure of only about two man-weeks on operational research. It was by using such tools as simulation that complicated transport requirements could be really efficiently met, suggested the speaker. This particular case demanded 24-hour movement, but since BRSL did not plan to go beyond two 10-hour shifts, the starting times for each vehicle had to be staggered to cover the full 24 hours. Apart from time allowed for routine maintenance, the plan had also to provide for relatively frequent major service periods because of the intensity of operation.
The job involved two interlinked operations, one taking longer than the other; the best answer to this proved to be for one vehicle to complete two full cycles while its counterpart completed one; at the change-over point there still had to be an allowance for delay, but if this was set at, say, 5 minutes • maximum, then direct interchange would sometimes be impossible, and a spare trailer would have to be provided. But since this was a full /empty change-over, it was no use having a spare empty trailer waiting when a full one was required, and vice versa. Extending the simulation provided the eventual answer that the system would work with only eight vehicles so long as an empty and a full spare trailer could be worked into it.
Mr. Fielding said simulation techniques had been found very valuable in making worthwhile savings in the equipment and staff seeded for a particular job, but it was always essential to allow a margin for "disaster elements" in any plan, based where possible on known snags.
Introducing the BRS consultancy service to his audience, he said it arose because increasingly detailed answers on complex distribution problems were often beyond the resources of a busy branch or district manager to provide. Often the consultants acted as communicators—putting the specialist knowledge at the disposal of someone with a problem in that field.
The BRSL consultancy tackled (a) "trouble-shooting" tasks within the organization; (b) advice on distribution problems for outside customers; and (c) designs for future large-scale BRSL developments, which might be termed a form of corporate planning.
Factories out of phase
The speaker gave an example of the first type of work, explaining how on a long-standing industrial contract BRSL was very concerned that the cost of the job was not being sufficiently covered, while the customer for his part was adamant that the equipment and facilities provided were adequate for the volume of traffic. This was a round-the-clock inter-factory service and it took a consultancy team six months to establish the real cause of the trouble, which was an out-of-phase situation between the two factories being serviced, and not in the transport operation itself. The eventual result was the formation of an entirely new depot to serve this customer.
Mr. Fielding said that this provided an example of the difficulty of convincing a customer that he simply could not afford the 100 per cent service level for which he was asking. It was much less costly to offer a 90 per cent service level than to offer a 95 per cent level, and very much less costly than offering, say, a 97 per cent level.
On the distribution front, Mr. Fielding felt that problems had arisen with vehicle scheduling by computer not because the system was fundamentally impracticable but because of present limitations on feeding in enough relevant variables. He was certain that this problem would very soon be overcome—to the point where it would eventually become unacceptably inefficient to schedule vehicles manually for any large-scale operation.
If, for example, one accepted that, in any one field of distribution, nine depots were the ideal number for serving all Britain, then each depot would require many vehicles; and when a man had to schedule more than about 50 vehicles a day his scheduling efficiency deteriorated rapidly. because the human brain simply could not cope with the problems. But, properly fed, a computer would find such quantities child's play.
Studies of really large-scale distribution, said Mr. Fielding, suggested that the economies of scale were very relevant. An extrapolation revealed that the unit cost of the local delivery function continued to fall almost indefinitely; the administration cost dropped even more rapidly. He agreed, however, with a questioner that all these economies could come to nothing if management and supervisory staffs and methods were not developed in a way which kept labour happy and efficient in such huge establishments.
The speaker deplored the all-too-frequent practice of making payrolls the first task for a computer. This was, he said, the most costly way of dealing with payrolls. He would far rather see computers used for tasks too complex for manual solution. '
For example, if only it were possible to put the running costs of individual vehicles on to computer, it would be possible to determine the most economic replacement life of each vehicle on the basis of its own record. Compared with present generalizations about vehicle replacement, he felt this could offer tremendous savings, though he agreed that it would be costly and complex to extract and store the basic information.
The work-study (or, more properly, industrial engineering) section of the consultancy team was now heavily engaged on research for the British Road Services productivity scheme, said the speaker. Like other schemes, the object was to provide a bonus for higher output; early investigations had revealed "a remarkably low level of driver productivity", for many reasons.
Mr. Fielding said he was surprised that the people who actually controlled the traffic work had had no information showing the overall cost or revenue of particular jobs. This was now being altered so that traffic staff would be able to assess the profitability of a job. The aids to this would include a standard means of assessing the cost of multi-drop delivery work.
One of the important needs, said Mr_ Fielding, was a costing of traffic flows. A guidance system for this was being evolved. He agreed with a questioner that it was essential for each branch to know where its strengths and weaknesses lay, so that one customer was not subsidizing another.