It also contradicts what the other commenter said about barcodes originating with railways. But I believe Derek.

It looks like the video touches on the origins of product barcodes (UPC/EAN) specifically, not literal barcodes in general.

Edit: Oh, I guess that's moot since the product barcodes predated the train barcodes anyway.

Kartrak is generally recognized as the first barcode system in actual use, but research on the topic and small prototype systems go back around 20 years earlier at IBM. So it depends on what you consider "first."

One of the issues is that barcode readers were very large and expensive to construct in the mid-century. The railroad application became practical earlier because of the small number of readers and large amount of space available at railyards. Kartrak readers were small-refrigerator-sized cabinets with arc lamps and an AC motor driving a rotating mirror. They required regular mechanical maintenance. The actual logic was done in a minicomputer installed in a nearby building. Between the optical cabinet and the minicomputer, it probably came out to something like 15 square feet for each reader (and kW power consumption for the arc lamp).

Practical retail barcodes had to wait for a lot more miniaturization of the mechanics and pretty much for lasers, and weren't seen until after Kartrak.

You can tell that Kartrak had a rather distinct lineage from retail barcodes - GTE, who designed Kartrak, don't seem to have been aware of the earlier work at IBM and designed their system independently. WABCO developed a competing system that didn't gain adoption but actually was based on the IBM work and resembles modern barcodes much more. The result is that Kartrak is an exceptionally weird symbology, with a number of design traits that were either not seen at all in other barcodes (the unusual half-toned bars for better performance with arc lamp readers) or not seen until decades later (the use of color and offset start/end points of bars to avoid partial reads).