Does anyone want an 8 foot kettle lead?
This story deals with an error caused
by humans. Computers participated but they did well and even
demonstrated the error, albeit too late!
Several years ago when a particular large project was moving
from one company business to another it was relatively straightforward.
All the components needed for the manufacturing phase were coded
with company house-codes; all had been entered into the company-wide
computer databases correctly; and all the supporting documentation
had been completed.
At the precise stage in the transfer
of the project however the R&D arm of the company had already
started to input the final customers' requirement into the Automated
Purchasing System and this was taken over and completed by the
in-house manufacturing organisation chosen to build the equipment.
Although this was about 80 miles away, thanks to telephone lines,
the computers in both places knew exactly what was wanted and
over a period of about 9 months or so
most bits had arrived
in the stores of the manufacturing plant.
Regular meetings were held and progress was monitored. Budgets
were closely checked and discrepancies having the smallest deviation
argued at length.
Things were more or less on schedule
was noted that some things were arriving a little later than
planned but contingencies were in place to overcome this.
Week followed week and real progress
became apparent as large trailers with metal containers fitted
to them appeared in rows and gradually filled up with equipment.
More and more of the factory began to fill up with these dark
green containers and people began to relax. Finances were just
within budget and at least this bit of the contract should make
At the end of one memorable progress meeting I recall asking
if a couple of spare "kettle leads" were available
for the test rig back at the R&D establishment where testing
was in progress (these so-called kettle leads were used to interconnect
radio equipment to the mains supply and had been specially designed
for the project, looking like moulded kettle leads but being
about two or three times their length).
The Manufacturing Director stepped forward.
"Allan," he said, "I've
just been learning about the intricasies of the "Bill of
Materials" computer system and I'm now a whiz with a terminal.
Let me help". So saying he went to the nearest computer
screen and started typing in the code I supplied for a kettle
"There!" he said triumphantly,
stepping back to admire his handiwork
"You can have
any number between 2 and 9000".
I said 2 would be fine but why did we
have 9000 still in stores?
Silence descended on the remaining members
of the meeting who were just chatting, drinking coffee, and generally
packing their papers into briefcases etc. prior to departing.
Some rushed away.
What had happened was this.
When the contract details had first
been computerised we were unsure of the numbers wanted of one
particular set of equipments. Some said it was 30, others said
29 and some said 31. No-one was really sure.
The piece of paper carrying the information
for ONE set was an Items List. The automation at the R&D
establishment wasn't as complete as it could have been and quantities
of major equipments were manually entered from the contract documents.
The Items List had been supplied to the Purchasing Clerk, who
co-incidentally was a young girl who had only recently joined
Across the top of the original copy
of the Items List had been written in blue biro "x30".
The figure had been crossed out and "x31" had been
entered. As everyone knows when you copied documents written
in blue biro, results were not always consistent. In this case
(it was later discovered) the "x30" and "x31"
had come out OK but the line crossing out the lower figure had
not. The young girl, working with a Xerox copy, had duly entered
"930" (ie.30 x 31).
Each list had a quantity of 10 kettle
leads hence we ended up with 9,300. We only needed 300 so we
had 9,000 too many.
So no big deal you might say. Sell them.
First, we couldn't because kettle leads
(at that time) couldn't be longer than a prescribed length which
was less than half of ours. Second, what about the rest of the
items on the Items List?
Well put it this way
when we tried
to sell surplus Military Connectors to other companies we were
met with initial comments such as "yes please they're as
scarce as hen's teeth .. for some reason we can't get any"
. It turned out that British Industry was having a mini-boom.
All our suppliers were making things for us by the tens of thousands!
Connectors were delivered by the van load for months.
Unfortunately our customer had wanted
everything in a special matt black finish. When we quoted codes
to would-be buyers they queried the slash MB at the end. "Oh
that's no good, we only buy in NATO green" was the universal
comment. The matt black finish rendered the surplus scrap!
There were about 35 items on the Items
List in quantities of up to 30 (ie. the total of some bits was
27,900!) and the cost of the excess components turned out to
be £2,000,000. I believe a local scrap dealer came in and
bought the lot for a couple of hundred quid. At future progress
meetings, when it came to manufacturing spend against budget,
comments were somewhat muted.