The fascinating history of barcodes – Part 1

In 1948 Bernard Silver, a student at the Drexel Institute of Technology (Philadelphia) overheard a discussion between the president of the local food chain and one of the deans about researching a system which would enable them to automatically read product info at checkout – a system we now of course cannot imagine life without. Bernard Silver and his friend Norman Woodland were fascinated by this challenge and began to work on a variety of different systems.Their initial system – which used ultraviolet ink – actually worked; but the price of the ink and its tendency to fade was problematic. Undeterred, Woodland left Drexel and began work on a system inspired by Morse Code. His very first barcode was formed with sand on the beach!  He extended the dots and dashes of the code downwards to form lines, and eventually read them by shining light from a movie projector through paper. Trying a different approach to allow scanning from any direction, he decided that a code printed as a circle would actually work better, and in 1949 he and Silver filed a patent application for what they named “Classifying Apparatus and Method”. The initial bar code reader which Woodland and Silver constructed in Woodland’s living room was the size of a desk and required wrapping in black oilcloth to keep light out!

The patent was issued in 1952, and Woodland, now working for IBM, continually tried to interest them in developing the system. Eventually they made an offer for the patent, but were outbidded by a company named Philco – this 10 years later in 1962, and it was eventually sold to RCA. (Not the record label by the way – the company eventually to be taken over by General Electric.) Tragically, Silver died the very next year at age 38.In the meantime, David Collins – who received his master’s degree in 1959 from MIT – was working on a system to automatically identify railroad cars. He developed a system called KarTrak for Sylvania Corporation, which used coloured reflective stripes in blue and yellow attached to the side of the cars. Light reflected off the stripes and was fed into photomultipliers. (Basically a vacuum tube which is an extremely sensitive light detector). The system used a 6-digit company identifier and a 4-digit car number. Tests began on gravel cars in 1961 and continued until 1967 when it was selected as the standard identification method for the entire North American Fleet. Unfortunately the system was easily fooled by dirt and was abandoned in the late 1970’s, before implementing a similar system based on radio tags in the 1980’s.

Although the initial system had failed, requests for similar technique were received from a toll bridge looking to scan for cars which had purchased monthly passes; as well as from the US Post Office to track their trucks and even for a pet food company. This, of course, sparked interest in the grocery industry.

In 1967, Collins went to Sylvania’s management looking for funding for to develop a black-and-white version of the code for other industries. He recalls: “I said what we’d like to do now is develop the little black-and-white equivalent for conveyor control and everything else that moves”. They refused, saying that the railway project was more than large enough for them and they saw no need to branch out so soon. In retrospect, clearly not a wise move!


>>Read Part Two Here<<