Mirrored Bars

Barcodes and Colors
I would bring my copy of the Distribution Codes Inc three ring binder into the Red Briefing Room and tell the guests that what we were going to do is help them understand what was behind some of the information in it. I’d tell them how the bars and light spaces were viewed in the scanner’s logic. In fact it was really almost impossible to misread a symbol even if there were some imperfections. Because of the cross checking logic embedded in the technology and the POS system, even severely misprinted symbols would only fail to read, not misread to a wrong value. Most of the time the technology was so tolerant even bars missing up to 20% of their width would commonly read correctly.

But there was a relatively small paragraph at the top of a page near the back of the manual that discussed color. It approximately said, “the supermarket scanner is a helium-neon laser which operates in the red frequency range.” This means that the laser will see red as white and see any colors with no red in them as if they were black. The notebook recommends that packaging people obtain a Kodak Wratten #26 filter which when placed over the symbol allows a human to see the symbol the same way a laser sees it. IBM believed that if the symbol was visible looking through a Wratten #26 filter, then the checkout system would also see it. One of the first symbols picked up to demonstrate the scanner was a pack of Wrigley’s Spearmint Gum which had black bars on a green background. It looks completely dark to the scanner and under a Wratten #26 filter, so it couldn’t be read. We also got red bars printed on a nearly white milk carton. Red on white looks all white to the scanner and therefore can not be read. In both these examples the package appears uniformly colored, there is no differentiation for the scanner to interpret as bars and spaces.

In the fourth session this got a lot of attention. As I spoke to this point I heard an unfaltering low-voice come from out from the second row just to the left of me in the group of 30 or so guests, “Coca-Cola’s colors are red and white!”

The word Coca-Cola immediately catches one’s attention. I quickly responded as politely as I could, “Sir, please understand, this isn’t IBM talking or anything political in the industry. This is optical physics!”

“Coca-Cola’s colors are red and white!” came back the patient low-keyed response.

Trying again I softly added, “Please understand, this is Mother Nature, not any company or organization!”

And again, very patiently, came the response, “Coca-Cola’s colors are red and white!”

I think I was beginning to understand! So about 10:45 am during the first demonstration when we let the guests go into the lab and handle the equipment, I got one of the other marketing people to help Jan Mosser with questions and I hustled up the hall in Building 602 to find Joe Woodland. I told him the situation which he grasped immediately. He indicated he’d get back as soon as he could. Joe then went and talked with our optical expert - Lee Dixon, and George Laurer to consider how to resolve the issue. The grocery manufacturer briefing session went ahead on its regular schedule, but about 2:30 pm Joe Woodland came back and asked to make a proposal to the group.

I believe Lee Dixon had come up with a solution. Joe described how the scanner only sees the reflected light. So if the surface is very shiny like a mirror or aluminum can would be and it also is a curved surface, it reflects the whole beam off, still as a coherent beam, in some other direction than towards the Photo Multiplier Tube reading device, the device that collects the light. There would be no diffused light some of which might be picked up by the light collector. This is equivalent to absorbing all the light and would look to the scanner like it was a dark bar. When the beam crosses the shiny part, it puts no light back into the light collector so it looks black to the scanner system. Painted and paper surfaces scatter light in a diffuse pattern some part of which does get back into the scanner’s PMT, Photo Multiplier Tube. But, unpainted shinny surfaces on the can reflect all of the laser beam’s light away in some direction other than where the light is collected, i.e. it looks like the black bar.
Red and White Coco-Cola package with symbol mark solution. The unpainted bars were highly reflective and therefore appeared as a black bar to the scanner.
That was the birth of using unprinted shiny areas to represent the bars on the aluminum cans. We had found a way for Coca-Cola to source mark their soda cans and still only print red and white. Their representative at the meeting thought that might be acceptable.