Blockchains are genuinely pretty promising for supply chain tracking. Distributed ledgers are most famously useful when you distrust the motives of a central au

Blockchains are genuinely pretty promising for supply chain tracking. Distributed ledgers are most famously useful when you distrust the motives of a central authority like Bitcoin, but they're also a reasonable option if you distrust the accuracy of a a central administrator. So for the purposes of supply tracking, different Coca-Cola facilities and shipments are 'individuals' which might report mistaken or dishonest results to the central server. And once somebody screws up, that trusted authority becomes a problem for others facilities to work around. For sufficiently large and restrictive systems (like US military supplies), correcting an error can become functionally impossible. At that point, you start resorting to awful two-wrongs-make-a-right solutions like entering fictional shipments which "move" a misdirected item from the listed location to the real one, or even redoing needless part replacements to match reality to documentation. Obviously you can track supplies without a blockchain, and I'm sure a lot of Coke's actual gains came from tearing out a bad system and replacing it, but a blockchain does at least encourage good tracking design (one authoritative record per item, transactions are assessed by peers rather than immediately accepted by an authority). And if the goods in question have individual identifiers, "proof of work" is actually a great addition. It doesn't have to be computationally hard if you trust all the users, but you still get a system where "I am holding this and hitting it with an RFID scanner" is allowed to overrule any number of past errors regarding that object. (Did Coke get all those gains? No idea. They probably just scrapped some legacy nonsense for a not-too-stupidly designed system.)