a thoughtful web.
Good ideas and conversation. No ads, no tracking.   Login or Take a Tour!
comment by ButterflyEffect
ButterflyEffect  ·  1982 days ago  ·  link  ·    ·  parent  ·  post: Pubski: October 24, 2018

More relevant. I'm in a food safety conference and the concept of utilizing blockchain technologies (Ethereum) for food supply chain traceability. insomniasexx or mk. Thoughts?

Apparently a group was able to use QR Codes on bottled wine which traced back to temperature and environmental conditions at the specific vineyard.

Edit: somehow we've talked about crypto kitties and IBM Food Trust in the same paragraph.





kleinbl00  ·  1982 days ago  ·  link  ·  

The heavyweights are welcome to elaborate/contradict but the fundamental innovation of blockchain technology is its trustlessness. Know how much of your world is dependent on ISO 900x as governed by TuV or the equivalent? Inject "blockchain" into it and properly implemented, TuV is removed from the equation.

A real-world case example would be, say, insulin shipments. Maybe you can go down to CVS and buy your insulin for $250 a vial. Maybe you can call 1-800-DIABEETUS and get it for $200 a vial. Or you can get on the Internet and find an outfit called "Ashraf's Insulin & Kidney Korner" that has it for $100 a vial. But when you look into it, Ashraf is in Peshawar and insulin is sensitive to maltreatment in shipping.

The cost of buying Ashraf's Insulin will go up if you need to verify that it made it from Peshawar without hitting degrading heat along the way. But you could take one of these, get a little IoT on it and equip it with a smart contract that will only release the buyer's escrowed payment to the shipper and the seller if it makes it to the buyer's GPS without the shock and temperature values being exceeded. You could also escrow the buyer's deposit on the tracker until it gets scanned at a return depository intact. And all of a sudden you're buying Pakistani insulan at half off without having to trust (or pay extra to) Fedex, Paypal, American Express or anybody else. Realistically this means that anybody with a bike or a piper cub can contribute to getting you your insulin; it also means that anybody along the way can view the data you're collecting (assuming it's a public rather than private blockchain). Put a barcode on the box, use that magic HTC blockchain phone and anybody who wants a piece of the shipping fee undertakes the liability of the shipment for as long as its in his care and pays him a portion of the overall fee depending on how quickly and safely he gets it x much closer to its destination.

You can use a QR code to trace wine bottles without invoking the blockchain. It's kind of the fundamental backbone of the logistics industry and how they know that these other 10,000 heads of lettuce need to be pulled because these other two heads of lettuce sent someone to the hospital. Juicero was leveraging that trick; that was their justification for $8 bagfruit. If it's expired (based on the QR code) the juicer won't squeeze it. If it's been pulled for food safety (based on the QR code) the juicer won't squeeze it. Typically those codes are per-box instead of per-bottle but they could for sure be on every single bottle. Put an RFID code on it while you're at it and you can siphon off the logistics every time it's within range of a connected reader from the vintner's cellar to the supermarket's door scanners.

ButterflyEffect  ·  1982 days ago  ·  link  ·  

I don't disagree with your example (while ISO 900x is prevalent there are at least two other, generally equivalent and widely used, food safety and quality certifications out there). And yes, RFIDs are great and I personally wish were used a bit more extensively. What was talked about today is the potential to greatly reduce recall event time lines, FDA investigations, etc. The issue was described as:

1. The only FDA traceability requirement being 1 up and 1 down in the supply chain from the manufacturer. Which leads to 2.

2. The lack of interoperable data from farm to retail store. If there is a lot of lettuce with E Coli from one farm, which went to five ready to eat manufacturers and eight produce packers, all of whom have different lot code designations, and all of whom ship to numerous distributors who in turn ship to even more numerous retail outlets, you end up with a very complex and time consuming traceability process.

kleinbl00  ·  1982 days ago  ·  link  ·  

I've never done food safety. I've done FDA and general manufacturing and TuV rules that world so it was an easy example.

So if I'm understanding the discussion, your point (1) issue is that the granularity of tracking as it currently stands ain't great and your point (2) issue is that what tracking is available it's proprietary, not interoperable and cannot be parsed by an external observer, correct?

So the first way to deal with this is through monopoly. Amazon decides to sell insulin. Amazon's Choice Insulin is made wherever Amazon feels like making it, and it's tracked from dead pig to doorstep using Amazon's logistics. Amazon owns the farms, owns the labs, owns the storage facilities and owns the trucks.

The next solution is through a large logistics company. This has been UPS' approach for decades: "we got great logistics and we'll let you buy into it." They may not make the insulin or sell it but the more deeply involved they are in your supply chain the better your insight into your supply chain.

Let's make an example of Chipotle. Chipotle ends up with more food safety scares than most fast or fast casual chains because of their commitment to local. This means they have a diverse and non-standard logistics chain which gives them crappy insight into what came from where. Compare and contrast with McDonald's - more than half of their french fries worldwide come direct from Simplot. Nothing wrong with Simplot? Nothing wrong with McD's french fries.

The useful aspect of blockchain is it gives you Simplot-grade trust, UPS-grade tracking and Chipotle-grade diversity because you're replacing UPS with a trustless organization that anybody can interact with. Any vendor under smart contract is subject to Simplot-grade homogeneity without Simplot-grade oversight. And Joes Tacos can get massive agribusiness-grade supply from anyone who is willing to interact with the smart contracts you wish to govern your logistics chain with.

The ultimate goal of blockchains is disintermediation. Right now you need an organization like UPS to provide you with traceable logistics. You need a Simplot-sized vendor to give you supply chain homogeneity. And you need Amazon-sized leverage to demand adherence to your standards as a manufacturer. But with blockchain, the contracts are up where they can be read by anyone and the standards to adhere to are available to all. Because the blockchain is trustless (self-reinforcing), nobody gets paid unless they stick to your standards and you can't stiff anyone if they adhere to the contract. Bob's Blockchain Business might wrap all this up and hire a bunch of guys to be logistics to compete for these contracts while taking 5% off the top; if he's on a public blockchain, Charlie's Crypto Commuters can take that same organization and do it for 4% without getting out of bed. The end result is that anyone nimble enough to meet the requirements can provide the same level of trust as a Fortune 500 logistics company, and anybody on the blockchain can experience the same level of service as a Fortune 500 vendor.

Make sense?

I don't know if I've recommended this book to you before, but if so, I'll do it again. It's "what the fuck is a blockchain" for normies that understand how to balance a checkbook but not compile C.

goobster  ·  1981 days ago  ·  link  ·  

There is a new competitor (called Samsara) in my company's field (GPS tracking for fleet vehicles), and they are taking the IoT approach to the problem.

Big picture: they don't really care about the vehicle itself. What they want to do is put a sensor on that object when it is made, and track all of that object's key metrics throughout its life. From the factory, to shipping, to inside the truck, to installation at the end customer's location, to disposal.

All that data goes up into the cloud, and they do deep and broad analytics on all of it.

It scares me a little bit when a key competitor doesn't even notice me. We track vehicle location and performance. That's just one small part of the whole chain-of-custody that Samsara is tracking. We aren't even a speedbump to them. (And they have $80m of VC money, and no need to make money... only acquire customers, so the VCs can sell the biz and make their 10x on their initial investment.)

Blockchain could be used for this... if everyone in the line of custody has it implemented and integrated systems to handle contract hand-off.

But it is a tough sell. Samsara already gives their hardware away for free, basically as a marketing expense. What they want is that sweet sweet monthly SaaS revenue. $x.xx/month/sensor adds up quick.

For them to convert from the easy-to-sell IoT pitch to Blockchain, they need to sell a lot of really traditional businesses on revamping and modernizing their internal systems. (Trucking, manufacturing, warehousing, etc.)

With IoT, the intermediaries don't give a shit if your box is sending out radio signals into the ether. They just pick up the box, move it to a new location, and put it down again. Simple, easy, and cheap on manpower.

I WANT blockchain to take over. But it's a big lift in many entrenched industries working on small margins, and who are generally adverse to technology-based solutions.