Losing count

Today we completed our physical inventory (first count) in record time. Tomorrow we will find out if we completed it with record accuracy or not.

Our inventory control is gradually improving–I think it is. I didn’t run into any pallets where every half the boxes were opened with some indeterminate quantity removed, or any area that basically amounts to a gigantic industrial-sized junk closet. But the ideal of having an actual count for every item still remains a dream.

Some parts weigh in the thousandths of a pound. Actually I think some were in the ten-thousandths. Some weigh several pounds a piece. You can scale count the smaller parts but you have to get a large enough sample size to meet the sensitivity of the scale counted out by hand.

A stray staple can throw off a part count. A breeze on the scale can skew it. The plastic bag holding the parts can weigh as tens of parts, easily; if the count of pieces inside the bag is not marked (and how knows how accurately) by the supplier, we have to come up with some way of identifying the number.

Take a roll of stickers, for example. I personally weighed a complete roll of stickers that I was told numbered five hundred, and then another partial roll that came up as 155. But this certainly gave the wrong count of stickers. The equation would be:

R + (W*X) = Wx
R + (W*Y) = Wy

where R is the weight of the cardboard tube, W is the weight of 1 sticker, X is the count on the first roll and Y is the count on the second roll. Wx is the total weight of the first roll and Wy is the weight of the second roll. But the method we actually used to count assumes:
W*X = Wx
W*Y = Wy

Does it really matter? Shouldn’t we just keep enough buffer on hand of these tiny parts so that ten, fifty, or even 200 piece discrepencies don’t matter? Well, we are not going to be 100% accurate so we do need to keep a buffer stock. But without any good way to keep track of where the actual count is headed, all you do is gradually eat up your buffer without knowing it. And then you shut down your production for lack of an item that costs fractions of a penny per piece. It’s happened. More than once.

After we finalize this inventory count we are supposed to begin a cycle-count program on all stock to help us better track our counts. This may help, especially if it drives us to improve our control so that I can’t pretty much pick any 10 square feet of the assembly area and pick up a half dozen parts off the floor from under benches and racks.