4 ms·
The supply chain argument for blockchain basically goes like this: 1. supply chains tend to require lots of coordination, but, also tend to be made of entities
by abakker 8y ago
The supply chain argument for blockchain basically goes like this:
1. supply chains tend to require lots of coordination, but, also tend to be made of entities that are each individually maximizing for value.
2. In the past, this meant that they needed to coordinate commercially, but only just enough that the supply chain met everyone's bare minimum, contractually-defined needs.
3. typically, this mean that systems were not well-integrated, and the data often didn't make it more than one or two steps in the chain before it fell off.
The problem is that everyone in the supply chain knows that better data would improve their operations, but nobody is willing to cede control of a centralized system so that everyone's data can been coordinated and managed. The semi-adversarial relationship that exists due to commercial terms means that no member of a supply chain wants to give up more data than they have to to a central entity.
Alternatively, the problem can also be viewed as a dimensions problem. In reality, there is no natural "owner" of a supply chain. pick any member of the supply chain, and put it at the center, and the data that it needs, and collects will be a different set of characteristics than all the others. there is obviously useful data that spans the whole supply chain, but for each member, the way that data is grouped, pivoted, counted, and accounted for can be different.
Perfect data entry into a system could solve the latter problems, and coupled with differential privacy or other masking functions, you can enter lots of a data into a system and each member of the supply chain should be able to manage their business better, which in turn improves the performance of the whole system.
The blockchain component is really only for the function of solving the system-ownership prisoner's dilemma. Effectively, each member of the supply chain owns the system, but its decentralized nature means that also no single member owns it. A blockchain helps create a system for managing a supply chain when there isn't an owner of the supply chain.
- hef19898 8y agoIn every supply chain there is a owner, the channel master. Basically the guy that controls it, directly and the farther out it get indirectly. Until a new party is the channel master for that part of the network. Besides the issues of matching goods to information the transparency that is necessary for a true fully integrated supply chain as envisioned or necessary for blockchain to work won't be accepted by any player. Si ply because everything from markets to customers to potential BoMs and production efficiency can be derived by any party in the chain. And that is not even without looking at processes which is different story all together.
- aey 8y agoI don’t see a specific need of a “blockchain” feature for this usecase. Uber didn’t need decentralization to coordinate drivers and clients. Aws/Postgres can handle the fault tolerance. FedEx already does this.
- abakker 8y agoI agree, there is no need for a blockchain. But, that explanation is what has been proposed as the justification for using one. I would counter also that neither of your examples are actually supply chains. FedEx is part of a supply chain, but only a single member. the chain includes the manufacturer, wholesaler, etc. Uber also doesn't have a supply chain with its drivers. It is not running a supply chain for that purpose, it is a marketplace.
- Canada 8y agoYeah I agree that blockchains look attractive for handling the ownership problem, but I don't believe they are necessary for cases like this. A well designed open protocol would do the job just as well. Blockchains are for cases where double spending needs to be prevented. It seems to me that logictics merely needs to broadcast assertions by the participants. Kind of like certificate transparency.