In this age of data science and big data analytics, industrial engineers have to know that their early scholars promoted use of data for decision making through records maintained over large number of years. Taylor's study of belt designs was based on years of data accumulated on the consumption of belts and the number of times belts were to be tightened and replaced. Harrington Emerson wrote a full chapter of the topic of maintaining records.
(
Harrington Emerson - The Twelve Principles of Efficiency)
WHEN a child touches the red-hot end of a poker, the information, advice, notice, record is reliable and lasting, also immediate and adequate. The scar is a perennial reminder of the mistake. Many of Nature's warnings are reliable, immediate, and permanent; they reach us and other animals through the senses — we hear, we see, we smell, we taste, above all principally, we feel. There are two nerves from the brain to the eyes, two to the ears, two to the nose, two to the palate; there are several hundred between body surface and brain. Very few people allow themselves to be burned, because the penalty is reliable, immediate, and adequate; but they are not as shy about more deadly disease germs (probably a thousand people die of tuberculosis for one who is burned to death) because the result is not reliable nor immediate.
The object of records is to increase the scope and number of warnings, to give us more information than is usually received immediately through our senses. A steam boiler with water in it, a fire under it, and all outlets closed, is more dangerous than a hot poker. There is very little to indicate the imminence of disaster. It is too hot to touch with the hand, although it is conceivable that a spot in it might be so insulated as to permit the engineer to tell by feeling whether it was becoming too warm. A thermometer would give a better record; but usually there are three recording instruments, each reliable and immediate, one of them in addition adequate. The engineer watches his pressure gauge, he watches his water-level glass, and the safety valve will pop even if he has fallen asleep. It is because of these three devices, one of which is independent of the man, that there are so few boiler explosions. All around us are many natural forms of advice, of records — the word is throughout used in its largest sense.
The object of records is to annihilate time. To bring back the past, to look into the future, to annihilate space, to condense a whole rail-road system into a single line, to magnify the thousandth part of an inch to foot-rule measurement, to gauge the velocity of a distant star by the shifting of the lines in the spectroscope, to annihilate temperature by enabling us to read the millionth of degree or the 10,000-degree difference between moon and sun heat.
Animals make and use records, reach out to each other through time and space; and the naive surprise of the doe when the stag appears does as much credit to her modesty as the trail of musk left in her footsteps along many miles and for many days does credit to her involuntary common sense. Man alone reaches out to man through millenniums; and the pictures carved in stone, the hieroglyphics pressed in brick or cut in granite, tell us more about the intimate lives and philosophies of the Hittites, of the Egyptians, than we know of our own immediate ancestors, the Germans or the Gauls —than we know of our immediate neighbors, the Indians. Pictures and writing were a great invention; the reducing of music to written form so it could be reproduced was even more marvelous, since through the eye we recreate for the ears, thus bridging the gap between the senses. The perpetuation of sound through ages in the phonograph disk, the perpetuation of movement on a long film, these are part of man's triumph through records. The phonograph disk is, next to the brain, the most marvelous, if not the most useful, record man possesses, since all the throbs, moans, triumphs, all the nuances of a hundred instruments and of a hundred voices, pulsations of the air, are recorded by the needle point in a microscopic line; and that line, that perfect record, gives us again the same air pulsations, the same great instrumental and vocal chorus.
Records are anything that give information. Men have always felt the need of records, but they have not always known what they wanted nor how to secure them. In the great industrial plants one knows not whether to marvel most at the absence of reliable, immediate, and accurate records, or at the superabundance of permanent records, collected with painstaking and at great expense, but neither reliable, immediate, nor adequate. Even if the latter have all these qualities, there is often great duplication, and as a consequence we find an immense amount of accumulation of very little value, which has cost far more than it need. An example of duplication may be found in the coal records for locomotives. Expenses of operating locomotives are generally recorded per mile, but suddenly a parallel set will crop up showing miles run per ton of coal. It has not been unusual in a great corporation's records to find a great variety of monthly tabulations, and when inquiry is made it is finally unravelled that twenty years before some president wanted a certain set of records, that his successor wanted a different set, which were started in parallel, that a third and fourth incumbent added their requests, but the old tabulations continue to be made and painstaking clerks work their monotonous lives away in neat compilation that no one has looked at, much less used, for a decade.
When the tramp piled and repiled the same cord of wood first on one side of the yard, then on the other, he was working efficiently but to no purpose; and having the soul of an artist he finally rebelled.
A clerical force may be hard at work, but it may accomplish very little and in the larger acceptance of the word it is inefficient, even as a hard-working steam engine using 50 pounds of steam per horse-power hour is inefficient in spite of its diligent consumption of coal.
There are records of all kinds, many of them essential to our continued existence. There are in a much more limited way records of cost; and between the two extremes of universal records (as the swing of the earth in its sea-sons or the slow aging of every living and inanimate thing) on the one side, and cost records on the other, come records of efficiency, and these are what we particularly need in the present phase of industrial life. We have not yet learned to use to any great extent the conception of efficiency. We are interested in what eggs cost per dozen, not in the weight of each egg; we ask the price of coal per ton, but rarely know whether it contains 10,000 or 15,000 heat units per pound; we violently resist a demand for a 10 per cent increase in wages, but we tolerate a 50 per cent inefficiency in the worker. Not one in ten thousand knows even approximately the cost of food. Its price is known, but not its value, and if a curve of food values per pound should be drawn, and above each item its price, the line would look like the record of the seismograph during an earthquake, or the record of a magnetic needle during an eruption on the sun.
The whole United States was frantic in 1896 over the money question, and not one in a thousand of the gold advocates knew that owing to violent fluctuations in supply and use gold had varied in value more than any other staple, not from hour to hour, as gold bonds and gold stocks fluctuate in value on the stock exchange, but from decade to decade. One of the tasks of modern scientific management, of efficiency and standard-practice engineering —two names for the same ideals — is to convert efficiency records into cost records, since the language of costs is understood by all, the language of efficiency only by the few. It is, of course, generally true that costs will decline as efficiency increases, but this is not always so.
A jeweller may work with the same efficiency setting on one day a $2,500 diamond in a gold stickpin and the next day setting a $0.25 bit of glass in a brass pin. Costs have varied, but not efficiency. A Japanese miner may work for $0.20 a day and an Alaskan miner for $15.00 a day. Each may work with equal efficiency, but the cost is very different. On the other hand, a farmer, from the same field, planted to the same crop, plowed by the same man, team, and plow, raises increasing crops of the same grain ; but wages, land values, and the price of horse feed might also increase so that decreased cost will not always directly flow from increased efficiency.
In the refinement essential for the control of modern operations, it becomes increasingly necessary to state efficiencies even if we talk costs.
Efficiency of Labor and Oost of Locomotive Bepaftn.
1905 1906 1907 1908 1909
_______ __________ . •
As a contribution to the solution of this problem a universal formula of cost and efficiency has been evolved which has the further advantage of showing what records are really essential and necessary, what form they ought to take and what records are useless, confusing, and to be omitted. All the necessary reliable, immediate, adequate, and permanent records can be obtained and maintained for less expense than is usually incurred for misleading, delayed, inefficient, and ephemeral records.
The costs of modern operations consist of three elements. For instance, in a recent year it may have cost to operate all the railroads of the United States approximately:
For materials $ 524,000,000
For personal services 1,021,000,000
For interest, depreciation, and other cap-
ital charges 1,210,000,000
$2,755,000,000
Omitting millions, we can set up the formula :
Total cost = Material + Per. service + Invest, charges
2,755 = 524 + 1,021 + 1,210
C (actual) =M (actual) + S (actual) + I (actual)
Let us assume that extended investigations show very inefficient use of materials, very in-efficient use of personal services and also over-equipment, and that from a practical point of view it might be possible to accomplish the same general result with $370 of materials, $780 of personal service, and $600 of invest-ment charges. 41 The formula of standard cost then becomes:
* These figures are used only for illustration, not as the expres-sion of a conviction.
C M S I
(standard) = (standard) + (standard) 4- (standard)
$1,750 = $370 + $780 + $600
The efficiency of the whole operation is :
C stan dard $1,750 - r . _ x _ „ .
C actual 2755 =e>3,5 P er cent.=Total efficiency=E
The relation of standard cost to actual cost gives the efficiency. This can be applied to each
sub-part >:
Material cost standard $370 Material
Material cost actual $524 =70 - 6 %= efficiency
Labor cost standar d ___ $780 -$ 4 <* _ Service
Labor cost actual $l,02i " efficiency
Investment cost standard _ $600 =49 6r = Investment
Investment cost actual $1,210 ' efficiency.
Actual costs can next be stated in terms of standard cost and of efficiency: —
Total actual cost^ ^ 1 s £ ndard cost = gML° =$2 7
Total efficiency 63.5 * z >'°&
Total Standard cost Standard cost Standard cost
actual s of material . of service . of investment
cost Material efficy."'" Service efficy 'Invest, efficy.
Total actual cos^+^+g =f 2>766
If we know in advance the standard or theoretical costs, if we know the current efficiencies, we can predetermine actual costs. What we all desire is to make the industrial machine as efficient as possible, to bring efficiencies up to 100 per cent, and when we do this actual costs will be the same as theoretical costs. We must first attack the problem theoretically. We must have standards and we must have efficiencies. When a pump or steam engine is tested, by every means we ascertain ideals ; we then com-pare actualities with the ideals and we ascertain efficiencies. Similarly, in the great industrial problem we set up ideals, we measure against them actual performance, and we as-certain efficiencies, and as for pumps, and for steam engines, so also do we use these efficiencies to prophesy future costs.
When actual and ideal performances are both recorded the relation in one month will generally serve to predetermine efficiencies in the next month, the relation of one year to predetermine efficiencies in the next year.
The elementary formula is, however, wholly inadequate for a real determination of efficiencies and has in fact led to most serious misconceptions and consequent mistakes.
Reference has already been made to the folly of the man who buys coal by the ton without knowing whether it contains 10,000 or 15,000 heat units per pound, who scrutinizes the cost of personal service without knowing its quality, invests in new machinery without counting its hourly cost, or without being able to keep it busy.
The cost of materials depends on two factors, the quality and the price.
Material cost=Quantity of units at price per unit.
M c =Qm Pm
What is wanted is that QP shall be a mini-mum cost.
The usual impulse and plan is to attack the price, P. This does not work. It is almost impossible to lower price, yet maintain quality. There is a constant demand for better quality and the tendency of prices is upwards. In the last ten years railroad presidents would have had great difficulty in buying steel rails at
less than $28 a ton. Q, quality, is the important factor. There is almost no limit to the re-ductions that can be made in quantity. Let us take coal as an example. The ordinary industrial-plant furnace, boiler and engine, use five to seven pounds of coal per horse-power hour.
By buying better coal, better furnace, better boiler, better engine and better service, coal consumption can be reduced to two pounds, in some instances to one.
Efficiency of production of power as to material is raised from 14 to 40 per cent up to 100 per cent. The distribution of power may, how-ever, be very inefficient. Air, water, and steam pipes may leak, there may be seven voltage drops in electric transmission. For 100 horse power produced in power house only 80 may reach the places of use. There is usually great waste in the use of power ; lights burn, pumped water is wasted, steam blows through steam hammers, compressed air is used to ventilate rooms or blow the dust out of clothes. The efficiency of use is rarely above 70 per cent. As-suming the efficiency of production to be as high as 70 per cent, that of transmission as high as 80 per cent, that of use as high as 70 per cent, we have an end maximum efficiency of 39.2 per cent. If, as often happens, produc-tive efficiency is as low as 14 per cent (the air-brake pump uses about 200 pounds of steam per horse-power hour), if the efficiency of transmission is as low as 60 per cent (I have known power steam pipes to be laid unlagged through running brooks), if the efficiency of use is 30 per cent (cities where water is me-tered use only one-third as much as those where it is furnished without check as to quantity), then the end efficiency of 14 per cent production, 60 per cent transmission and 30 per cent use is only 2.52 per cent. It is not because of price, but because of the dependent sequence of inefficiencies in quantity that QP usually admits of such very great reduction.
Materials actua^-r^gjf — ^= —
JtSJv C/ tnq XVmp
If EE'E" is only 2.5, P st could be increased 40 times without adding to cost, but a compara-tively small increase in P st doubling it for in-stance, may be the easiest, quickest and most economical way of increasing EE'E" mq to 10. per cent, 40 per cent, or even 90 or 100 per cent, as the case may be.
Therefore, in the last generation railroad executives were willing to pay more for steel rails than for iron rails, fuel consumers are willing to pay more per ton for oil than for coal, bridge builders prefer expensive wire rope to cheap cast-iron, for in each case as quality goes up, quantity goes down much more rapidly. What is true of materials is equally true of personal service. Labor, like material,
consists of both quantity and quality. The quantity of labor is measured by time, its qual-ity by what it accomplishes. The formula for personal service becomes.
S=time in hours multiplied by wages per hour S=TW
When TW seems too high there is generally an insane desire on the part of those in control to reduce W. This is naturally resisted most strenuously by the wage earner. As in mate-rials, it is not the price of the unit per hour that counts, but the quantity used. Also as in materials, there are inefficiencies of initial
quantity, inefficiencies of distribution, and inefficiencies of use. Let us assume schedules of different rates of pay for different classes of workers. I have known industrial plants to engage 600 men when 300 would have been sufficient ; I have known 12 men to be assigned to a job that 2 men could have done. There is in-efficiency of initial quantity of 50 per cent to 17 per cent.
I have known men that ought to have been earning $6 a day, in reality earning only $3 because they were in the wrong place, paid $3 for work that a $1 a day boy could have per-formed better; I have known a $75 a day ex-pert to be kept busy on clerical work that could have been done better by an $18 a week clerk. These are examples of inefficiency of distribution, varying from 17 per cent down to 4 per
cent.
The inefficiencies of use are so tremendous that their cause has to be explained. Up to about a hundred years ago, with the exception of a few windmills, a few sailing ships, and a few cumbersome water wheels, all the work of the world was done by the muscular energy of man and animal. It was used fairly efficiently, often strenuously. I have been fortunate in seeing and experiencing personally much of what was formerly the rule, as the porterage of freight and supplies over the Chilcoot pass on men's backs, 100 pounds to the man, and the killing, by overwork, of 3,750 horses out of 3,780 in the awful strenuousness, but la-mentable inefficiency, of the White Pass pack trail in 1898.
The discovery that we could use coal, oil, gas, mountain water-powers as sources of energy has changed all civilization. In the United States alone we have per inhabitant twenty times as much energy available as when I was born. The man whose manual labor it would take for over 500 years to spade up a section of unbroken prairie land, is quite inclined to think that he is using his time very efficiently if with team and plow he breaks up 640 acres in four years, when in reality with suitable equipment, mechanical tractors and gang plows, it could be done in 36 hours.
The man who would take a week carving by hand a small frame, might pride himself on turning out one frame a day with foot power, when in reality with moulds and automatic machinery he could turn out one frame a minute.
If, as I have seen, a man using a shaper over-runs the necessary stroke three-fold, if the machine's speed is only 30 per cent of what it ought to be with modern steels, if his feed is a 1/64 inch instead of a 1/16, if he takes four cuts instead of two, then his end efficiency is only 1.25 per cent. Men have not yet realized that the ages of muscular effort are passed, that work can no longer be measured in man-power or foot-power, that we no longer want the man who can spade twice as much, the man of burden who can carry twice as much, the man who can break a horseshoe with his bare hands ; but we want the man on the bridge of an oil-fired steamer, we want the crew of an oil-fired locomotive, engineer on one side with hand on power-moved lever, fireman on other side with finger on oil valve ; we want the crew
of mechanical tractors and gang plows, each man directing and superintending the evolution of as much uncarnate energy as 2,000 mien could have evolved using man-incarnated energy.
Assuming as a possibility in inefficiency of labor a quantity of 50 per cent, of labor distribution of 17 per cent, of labor use of 1.25 per cent, we have an end efficiency of 1/5 of 1 per cent. I have seen worse happen than this, for sometimes the worker did nothing at all, at other times was busy on wholly unnecessary work. As a general average, efficiency of sup-ply of work is not over 90 per cent; efficiency of distribution, if fitness for the work is in-cluded, not over 60 per cent, and efficiency of use not over 70 per cent, giving an end effi-ciency of 37.8 per cent, shading off from this maximum to nothing.
As to service, therefore, as in materials, it is quality that ought to be improved by paying
a much higher price per unit. It is not more strenuousness that is wanted; it is more effi-
ciency with less effort. As T goes down, W must go up both relatively and directly. The
locomotive engineer is paid higher wages than the Chinese coolie, and as part of his daily life
he enjoys luxuries unknown to kings a genera-tion ago, still unknown to Chinamen. The
coolie carries 150 pounds 20 miles in a day ; the American locomotive engineer and the fire-
man haul 6,000 tons 60 miles a day. Piece rates are physiologically and equitably vicious
and wrong. They put a premium on harmful strenuousness, instead of standardizing condi-
tions and operations so that greater output will follow less effort, but higher efficiency per unit
of time ; they are based on the assumption that output is dependent on muscular energy as it
was in former ages, instead of being dependent on a steadily increasing quantity of uncarnate
energy, combined with a steadly increasing quantity of incarnate energy, both directed by
a steadily increasing intelligence.
T cannot indefinitely decrease, neither can W indefinitely increase, and experimentally we
must determine what combination of TW re-sults in minimum cost.
In the diagram on page 224, the vertical lines A, B, C, D, E are records of different men work-
ing on similar jobs but at different rates of speed. A, the slowest worker, takes 10 hours
to accomplish a task. His speed is that of a lame man only able or willing to walk a mile
and a half an hour. Nevertheless, although he may be wholly unfitted for the work and the
work not suited to him, he has to live, has prob-ably a family to support, and he is unwilling to
work for less than $0.30 an hour, and if he is wise, joins a union which will enforce this mini-
mum rate. A's standard expenses probably eat up 90 per cent of his earnings, or $0.27 per
hour, his profit above expenses being $0.03 per hour. B is a faster worker, able to walk 2.2
miles an hour. He is also given $0.30 an hour, but in view of his greater speed an extra pay-
ment of 6.6 per cent is added, making his hourly rate $0.32. His living expenses, as for the
other man, being $0.27, his net earnings or profits become $0.05 per hour as compared to
$0.03. He has increased his profits 66.6 per cent. The man C is one who can and does walk
at the rate of 3.3 miles an hour, a mile in 18 minutes. This man earns $0.32 in wages and
a bonus of 20 per cent, making his hourly earn-ings $0.38. His net profit above minimum liv-
ing cost of $0.27 is $0.11 an hour, or an in-crease above A in net profits of 267 per cent.
D is a man who can walk 4.5 miles an hour, or a mile in 13.3 minutes. This is fast walking,
but not as fast as is regularly kept up hour after hour and day after day on the Yukon if
the trail is good.
D earns $0.15 an hour above the employer's basic rate of $0.32, his profit is 400 per cent
more than that of A. This man's speed is the most economical both for the employer and for
himself. A speed greater than 4.5 miles an hour is more than the normal man ought to
keep up. E is an abnormally fast traveler, running at the rate of 5 miles an hour, the
Yukon average. His pay rises to $0.60 an hour, his profit to $0.33 an hour, the profit alone
being more than the wages earned by A or B. His profit is 1,000 per cent greater than that
of A.
E is a strenuous but not an efficient traveler. His work costs more than that of either D or
C, and he will break down if he long continues the pace. If greater speed is wanted the
method must be changed, not the strain in-creased.
... . . Tst W«t
Actual service cost= B , t E n t B m w B mi w
W must increase as E t increases, W must fall as E t falls. If this is not the law, then
there is no hope ahead, and civilization, discov-ery, and appropriation of the energies in the
universe are disasters. But it is the law. Let us illustrate by a single example. Sixty years
ago $5 of free gold to the ton, $100 of combined gold to the ton, were about the lowest amounts
that it was profitable to work.
The average rate of wages for white men was low. The time efficiencies of gold produc-
tion have been steadily improved, gravels are now profitably washed that contain as little as
$0.05 to the ton, ores are mined and smelted that contain as little as $5 to the ton. Gold
production has increased from $13,500,000, the average before 1848, to $400,000,000 per an-
num. White men's wages have doubled and 250,000 men are now employed instead of
12,500 as formerly. Those who made money from owning gold mines have invested it, de-
veloping other industries, creating still further demand for employment. Let us assume that
the gold producers of the world should unitedly demand a 2-hour day at the same wage per
hour, instead of the present 8-hour day, on the supposition, firstly, that they would thus pro-
vide work for four times as many men, and that a larger proportion of the output of the
mines would go to labor. The immediate effect would be the closing down of nine-tenths of the
gold mines of the world, 225,000 men would be thrown out of employment, other industries
would be curtailed, still further increasing the supply of labor. The 2-hour provision might
stand, but either wages would drop until low enough to make the reopening of the mines a
paying proposition, or increased efficiencies would have to be applied to mining so as to
increase the output fourfold per man-hour of work.
More than ever before would it be necessary to make motion studies and time determination and to set up standards of supply, of distribu-tion, of use as to every item of work. If wages per hour are arbitrarily increased, the increase can be safely provided for by increased effi-ciency, and in no other way. If efficiency is arbitrarily increased, wages will inevitably rise, or effort will diminish.
What is true of materials and personal serv-ice is equally true of investment charges. In-
vestment charges, like personal service, fall into time for any performance and the cost per
hour.
I = T'R
in which T' indicates time in hours and R cost per hour for capital charges.
If all the railroads of the United States are worth $14,000,000,000, it is evident that the an-
nual capital charge for interest, depreciation, insurance and taxes might be $1,000,000,000 —
that the actual capital charge per hour is $114,155. If, therefore, as a token of respect
to the memory of a dead president, all railroads should stop operations for 10 minutes at the
time of his funeral, the cost would be about $20,000 in decreased efficiency of R, but the of-
ficials would hasten to make it up by increasing the output of the subsequent hours, thereby
raising the efficiency of T.
As for materials and for service, so also we must determine which T' and R in combination
result in the least cost.
In pay for services, the natural law is that an increase ought to decrease time in larger
proportion, but in equipment it is very common to increase R unwisely and very greatly for a
less decrease in T'. The same law prevails for equipment as for materials and labor. Addi-
tions to equipment should decrease, not in-crease, costs.
Muscular energy, whether of man or animal, is available only a few hours a day, 8, 10, 12.
Uncarnate energy is available 24 hours a day. The machinery in paper mills, in glass plants,
works 24 hours a day ; an ocean steamer on the Pacific will throb steadily for twenty days, the
big generators at the world exposition in Chi-cago and in St. Louis ran for six months with-
out a stop, big pumping machinery at mines will work even longer without shutdown. There
is, therefore, double and treble investment charge in working equipment only 10 or 8 hours
a day.
This was bad enough, but there was a boom period after 1897 that owed its start to
the Yukon gold discoveries, to a European crop failure with abundant crops here, and that was
further stimulated by the sudden expenditure of one thousand million dollars in the Spanish
war. America suddenly resolved to scrap all its old equipment and modernize from top to
bottom. Every railroad rebuilt its main lines with new grades, easier curves, heavier rails
and ties, rebuilt its bridges, stations and ter-minals, rebuilt or replaced its locomotives and
cars, built new shops and equipped them with new tools. Every city rebuilt its business
blocks and its aristocratic residence section, every street-car line was rebuilt and re-
equipped. Infected by the general contagion, every industrial plant tried to increase its ca-
pacity. Paper mills doubled the width of the paper machines, thus doubling their capacity,
iron mills became tonnage-mad, textile mills increased their machines beyond the world's
output of textile fibres.
What are we going to do about it? There are three correctives, and only three. Existing
equipment will gradually wear out, the country will gradually grow, but during the period of
readjustment those plants that are inefficient will be crowded to the wall and prematurely
die. Not only are American plants subject to high equipment charges because running so
few hours a day, but even for the 8 or 10 or 12 or 24-hour day, they are over-equipped and
much of the machinery lies inactive.
We have again and again found that ma-chines were not in operation over half the time
of a 9-hour day. When in operation they were inefficient. It is not so long ago that a loco-
motive-tire lathe would be run 18, even 30 hours, to turn up a single pair of tires, work
that on the same machine ought not to take over 3 hours.
