a computer company, then an operating system, and then looking for an LP programme that was compatible. As IBM was the de facto computer of choice to run company accounts, IBM entered the feed formulation arena in about 1965 as an add-on to their accounting software for manufacturing businesses. Most companies wanted to source computers from a single supplier and IBM was the safe choice as the mantra was ‘no-one ever got fired for buying IBM’. In 1963 the MD of an Irish feed company produced the Qeleq


Mark I analogue computer at a cost of about £4,000, which allowed a maximum of 24 raw materials in a single specification. Turning the knob on a raw material would change the inclusion rate as displayed by a light spot on the inclusion rate meter, and the inclusion rates of the other raw materials would change to compensate, as would the dial showing the total ingredient cost . At a stroke, hand formulat ion by mechanical desktop calculator, slide-rule or log tables was made redundant, er rors reduced, and the process of feed formulation was reported as 40


Edgar Pye (standing) monitoring the Qeleq Mark II at W&J Pye Ltd, c1971. Copyright Angela Bathgate. Four ‘spare’ raw materials (modules) may be seen stored on the top of the Qeleq


times faster - a week’s work reduced to one hour. By 1970 the Qeleq Mark II was released which was expandable from a base level of 14 raw materials and 12 specifications and with additional sections 28 raw materials and 19 specifications and a printer5


. The larger option


weighed about 600kg. At about the same time, the Richardson Select- O-Weigh process control system was invented, so mills had the ability to reformulate and implement receipes on a more frequent basis. The early adoption of this new technology gave pioneers a significant competitive advantage. Qeleq computers (Mark I and II) were sold to 30 countries over the next decade, but computer hardware was developing so rapidly that Qeleq was overtaken by the digital revolution and fell into receivership in 1976. It is sad to record that most nutritionists with experience of the Qeleq feel that the analogue method had many advantages over the newer and faster digital systems. In the 1970s DEC computers became competitors to IBM and


were at the forefront for scientific applications, and launched their PDP-11 mini-computers running RT-11 and RSX operating systems. These multi-user mini-computers were physically smaller and much more affordable than IBM mainframes, and some companies like Format International chose to run their LP software on DEC machines. DEC held their competitive advantage for more than 10 years, particularly with the


IBM PC. Copyright Ruben de Rijcke


VAX and mini-VAX. Then the IBM Personal Computer was launched in 19816


, a game-changer, and the world was suddenly different. DEC


countered with the DEC Professional 350 a single-user machine initially with two 1.2MB floppy drives, then a single floppy and a 5MB hard disk. For the first time, the cost of the hardware was lower than the cost of the software, and this gave rise to debate in many application software board-rooms, as to whether the price of the software should be reduced. The IBM PC was changing the world; it was affordable, scalable, could link to IBM mainframes, and ultimately it was clonable. Format was forced to re-write and recompile its software for the 640k RAM IBM PC and was able to sell its software for both DEC or IBM. In the early 1980s, Format’s Single-Mix® to least cost a single diet, and a Multi-Mix®


in a mill using its raw material stock and contracts, took 8-12 hours. Nutritionists of a certain age will remember Multi-Mix®


recording line by


line, ‘*Infeasible* with cost x’ until it reached ‘Feasible with cost y’ when one could relax and wait for a few hours for the lowest iterative cost to be calculated. Most feed companies would set up their Multi-Mix®


problem


during the day, and run it overnight. At this time the results could only be printed on paper, so there were several hundred pages of fan-fold paper to sort out on the following day (hoping the sprocket-fed printer paper did not jam). The advent of the super-fast DEC VAX with virtual memory allowed the launch of Global-Mix®


where all the formulations


of a multiple-mill company could be optimised at the same time, using all the raw material stocks in the company, and allowing stocks to be transferred from one mill to another. In the late 1980s with faster chips, single diets took seconds to solve, and Multi-Mix®


took minutes with


the results now viewable on screen. Format International was the first company to launch Parametric


programming in the early 1980s with VBREAK, SBREAK, TBREAK and MBREAK. Conceptually, this was not new as parametrics had been proposed as early as 1954 by I Gass and TL Saaty in their paper ‘The Parametric Objective Function’7. In essence, price sensititivity is parametric analysis in its simplest form. However, Format took this to a new level by iterative reformulation and moving certain constraints in steps, so that the nutritionist or buyer could examine sub-optimal formulations that might offer advantages over the absolute least-cost. Parametric analysis in Multi-Mix®


is a particularly powerful tool. In 1987, Kansas State University ran 27 problems through a


well-known LP programme running on an IBM mainframe which resulted in 7 erroneous solutions (26%), and deemed the errors with this software ‘too large to accept as an accurate LP solution program’. In the years 1984 to 1989 one of the occasional duties of this author and colleague Mike Ablett was to benchmark Format International’s LP software against competitors; often the results were the same, sometimes different. Sometimes we discovered that the competitors gave a feasible solution which was not the least-cost, which we labelled as multi-blend – a feasible solution but not the least-cost; possibly


5.


The author is indebted to Angela Bathgate for providing information about the Qeleqs used at Pye Farm Feeds.


6. Promoted by a Charlie Chaplin look-a-like https://www.youtube.com/watch?v=kQT_YCBb9ao


7. In Journal of the Operations Research Society of America, II (August, 1954), pp 316-319. FEED COMPOUNDER NOVEMBER/DECEMBER 2020 PAGE 35


, typically took 10 seconds run to least cost all the diets


Page 1  |  Page 2  |  Page 3  |  Page 4  |  Page 5  |  Page 6  |  Page 7  |  Page 8  |  Page 9  |  Page 10  |  Page 11  |  Page 12  |  Page 13  |  Page 14  |  Page 15  |  Page 16  |  Page 17  |  Page 18  |  Page 19  |  Page 20  |  Page 21  |  Page 22  |  Page 23  |  Page 24  |  Page 25  |  Page 26  |  Page 27  |  Page 28  |  Page 29  |  Page 30  |  Page 31  |  Page 32  |  Page 33  |  Page 34  |  Page 35  |  Page 36  |  Page 37  |  Page 38  |  Page 39  |  Page 40  |  Page 41  |  Page 42  |  Page 43  |  Page 44  |  Page 45  |  Page 46  |  Page 47  |  Page 48  |  Page 49  |  Page 50  |  Page 51  |  Page 52  |  Page 53  |  Page 54  |  Page 55  |  Page 56  |  Page 57  |  Page 58  |  Page 59  |  Page 60  |  Page 61  |  Page 62  |  Page 63  |  Page 64  |  Page 65  |  Page 66  |  Page 67  |  Page 68