Nestle Experiments with Tracking Gerber Baby Food on the Blockchain (wsj.com) 96
Nestle SA is putting some of its Gerber baby food products on a food-tracking blockchain to test whether the technology can trace the fruits and vegetables that go into its purees and squeezable pouches. From a report: Nestle's effort is part of a wider food-industry exercise aimed at improving food recalls by using the technology behind bitcoin to trace a worldwide ingredient supply chain. Food recalls can diminish consumer confidence and lead to lost sales. News of tainted baby food hits an especially sensitive nerve -- stakes that in part prompted Nestle to choose a popular variety of its Gerber linefor its blockchain test, said Chris Tyas, global head of supply chain at the Swiss company. Nestle offers more than 2,000 brands, including Haagen-Dazs, Stouffer's and Poland Spring. Nestle also sees the move as a way to generate customer trust everyday and during recalls. "People want to know, quite rightly, where ingredients they give to their baby have come from," he said. "We wanted a product in which trust meant something."
why blockchain (Score:5, Informative)
Why in the blazes they would even think about using blockchain (other than some C*O critter not knowing what buzzwords mean)? You pay a large cost for making data processed by untrusted nodes non-repudiable. The company controls all its data processing, and even if its distributed, can use far cheaper ways to ensure lack of tampering once a piece of data is committed, such as a simple hash of a block sent home when the block itself sits in the local database.
Re: (Score:2)
Why in the blazes they would even think about using blockchain (other than some C*O critter not knowing what buzzwords mean)?
You answered your own question.
Re:why blockchain (Score:4, Funny)
Re: (Score:2)
It's probably not going to be to the un-wieldly level of drilling down to the exact tomato, but whatever product has a complex list of ingredients it would be enough to know what supplier and batch of something was used to the point of where it makes sense.
e.g. a truckload of tomatoes from Bob's Farms might be bad and grabbing a whole set of whatever products was in contact with that truck - right down to who was driving is capturable data, No need to truly drill further.
And why blockchain? Oh c'mon - that'
yes but why block chain (Score:2)
Why not a central database or even a set of distributed database servers that allow a federated query of where the tomato went on it's journey to the table?
Re: (Score:1)
because it is difficult to agree on one central authority to store all the data. it is also insecure to have one authority. blockchain is a sort o distributed database - it's just a superior solution that is standardised, without counterparty risk and entirely in the cloud. why design some new solution when these things have already been figured out? there are many blockchain companies working on global supply chains - some prefer more centralised solutions and some less.
Re: (Score:3, Interesting)
because it is difficult to agree on one central authority to store all the data. it is also insecure to have one authority. blockchain is a sort o distributed database - it's just a superior solution that is standardised, without counterparty risk and entirely in the cloud. why design some new solution when these things have already been figured out? there are many blockchain companies working on global supply chains - some prefer more centralised solutions and some less.
No. Stupid. Dumb. Stop. Shut the fuck up.
Text file.
Product code.
Serial.
Location.
Date.
Some form of newline character (LF, CR, or both - I don't care). Hell, make it XML for all I fucking care.
Sign it with a private key.
Let anyone in the world mirror it.
Let anyone in the world throw it into a proper DB if they want to actively work with it.
Want to provide an API to access it? HTTP GET. Or maybe host your own database and let the world connect to it.
Block chains solve ZERO problems.
Distributed block chains
Re: (Score:2)
You should see the argument I have with score voting proponents. The biggest one--and the most ludicrous--has "Blockchain" in his name. He insists that elections aren't about electing the candidate the voters most agree upon, but rather on electing the "best" candidate with the most "marginal utility" based on how many points voters give candidates--and that this is totally-immune to tactical voting because the best strategy is to vote honestly.
It's like arguing with a flat earther, or a UBI automation
Re: (Score:2)
hmmm... as a voting systems activist I understand both your opponents motivations and also agree with your conclusions. I'd be curious if you have ever written up any of your thoughts on this. I've been mulling voting systems for decades and my views have evolved on it. Range voting is nice if there was an honest way for people to emit some single valued utility but there isn't so it fails. If there were then you can see why they might have a point. Interestingly if you take the opposite view and ass
Re: (Score:2)
Range voting is nice if there was an honest way for people to emit some single valued utility but there isn't so it fails.
In decision-making methods, committees define features of alternatives, weight them, identify the best alternative for each feature and rank that as '10', then scale the others relatively (Kepner-Tregoe method). They add scores to find the best alternative overall. The committee debates the necessary features, their weights, and how well each alternative satisfies these needs.
In range voting, the motives of individual voters are not consensus. A voter may rank a candidate as 1.0, and another as 0.5; t
Re: (Score:2)
I agree. To clean up a possible misunderstanding. In Approval voting one can cast multiple "yes" votes. So there isn't an issue with vote splitting. While the individual may chafe a bit because they actually would like to express that they prefer Bernie over Hillary even though they would accept either, it's still a lesser dillemma than being forced to vote for just one. Furthermore in agreggate over many voters, the average of voter biases for bernie relative to hillary will be somewhat restored. So
Re: (Score:2)
by the way, in my opinion, instant run off voting is pathologically bad way to handle ranked preference because anytime there are 3 equal strength parties it tends to not elect the centrist party prefered by condorcet.
A lot of far-left progressives viciously hate centrists, so consider this a feature. I say we need stability; if you want to shift the country left, then do so--the nation is its people; shift them.
1) in multi-precinct voting reporting results is complex since you can't actually determine the outcome at the precint. Now in the information age this isn't that big a hassle since you can report the entire ballot card not just the outcome.
That's the summability problem, also the Condorcet paradox (which is really just Simpson's Paradox [wikipedia.org]). Someone can win the Condorcet vote at every precinct, yet summing them together gets you a different Condorcet winner--unless they won Condorcet by winning the simple majority vote at every precinct.
General
Farmer-coop-shipper-distributor-Gerber ... Custome (Score:4, Interesting)
An individual farmer takes their celery to a local company or co-op, who has contracts with a nationwide or regional distributor.
The local company hands the shipment over to a shipping company, which brings them to a distributor.
The distributor sells them to Nestle/Gerber, through another shipping company.
Nestle sends some of it to their Gerber plant across town, some of it to their Maggi soup factory, etc.
After making the food from the ingredients, Nestle sells the baby stew to a grocery wholesaler. Another shipper.
The grocery wholesaler sells it to a small local store chain.
The store chain sends some of it to the store on Broadway.
The customer purchases it.
There are two big advantages of a block chain vs a traditional database here. That's a lot of different companies involved, in including a few trucking companies. They don't all use the same Oracle database, especially not the local farmer. Block chain is designed for many different people to be able to use it, adding entries, without conflicting with other in any way.
Nestle, and the customer, want to know that the local produce buyer isn't being lazy and making up records at the end of each week or each month. Everyone, including purchasers, can see that the local produce buyer added their first entry shortly after farmer adds "sold lot #74728 to Des Moines Produce Buyers".
Re: (Score:3)
You make no sense.
Anyone in a block chain can manage their own database if they want to track that shit. It doesn't have to be a central database.
You can requiring signing for all messages in any form of database. You don't need block chain for message signing.
You can have distributed copies of anything you want.
Re: (Score:2)
$500 billion. Compare bar codes, UL listing (Score:2)
Walmart, Nestle, Dole, Unilever and the other companies who started this project buy over $500 billion of food every year. If you want to sell food, you just might want these companies as customers.
Then there are network effects. Suppose a broker sells avocados to both Nestle and Kraft. Because they sell to Nestle, they and their growers use the block chain. So Kraft is buying from a participating broker, even though Kraft doesn't care. After the next horsemeat scandal or e coli incident, Walmart says to K
Re: (Score:2)
The UPC example is not a very good one. UPC barcodes offered benefits to everyone and
Farmers care about Dole. A lot (Score:2)
> Nestle or the brokers will not have a very easy time convincing hundred's if not thousands of farms
If you're a farmer, who exactly do you think is going to buy your 3 million head of lettuce that are ready for harvest on a 100 acre farm, if not Dole, not the food processors like Nestle* or Unilever, and not the brokers? Do you think you're going to sell 3 million at head to consumers at the local farmers market this weekend?
* If you're an American, you might associate the Nestle name with chocolate. Ne
Re: (Score:2)
There are two big advantages of a block chain vs a traditional database here. That's a lot of different companies involved, in including a few trucking companies. They don't all use the same Oracle database, especially not the local farmer. Block chain is designed for many different people to be able to use it, adding entries, without conflicting with other in any way.
There are 14 standards. We've made a standard to unite them all.
See the problem?
Let me spell it out for you: They don't all use the same Oracle database, so you stood up another Oracle database so they can all use that.
Re: (Score:2)
Gerber does not control all of it's data processing for this as they do not own, nor have visibility, over the entire supply chain. They don't buy fruit from farmers, they buy from suppliers who buy from aggregates who by from trusts who by from farmers.
If you do some reading (I did after seeing this article) the Food Trust blockchain is a great idea. Ignoring that it's controlled by IBM, and you have to pay them to participate, it means a one stop trusted supply chain for food as more and more suppliers
Re: (Score:2)
" blockchain is no better than some federated database with an easy API to which others have written user-facing interfaces"
The "better" is that everyone's copy is guaranteed the same.
Re: (Score:2)
Because mauve has the most RAM.
how does blockchain helps in most situations? (Score:2)
I don't see how blockchain helps in most situations. It's great for untrusted parties but for Nestle as well as most other companies trying to deploy blockchain, it seems like a secure centralized database would have less overhead. What advantage does giving each tomato a blockchain address have over giving each tomato a serial number? What advantage does recording each step a tomato takes in a blockchain have over just recording each step in a centralized database?
Re: (Score:2)
A blockchain is a decentralized, untrusted, read-only (as long as one party doesn't own the majority of nodes) database basically.
You can for example, as Nestle, publish the blockchain and then all sorts of consumers and agencies can "join" this blockchain and as long as no single party owns a majority of the nodes, the database can only be written to and never be changed. So you can eg. write all sorts of details about any operation into the database, it gets distributed and then later on, everyone can agr
Re: (Score:2)
You can for example, as Nestle, publish the blockchain and then all sorts of consumers and agencies can "join" this blockchain and as long as no single party owns a majority of the nodes
I've thought of using my access to several billion IP addresses to join bitcoin as the majority shareholder and manipulate the network, just to make it go away when people realize you can do that. There's a legal problem: I can't figure out who "owns" bitcoin, and it is impossible to commit a Federal crime against a computer service with no owner.
Here, let me help (Score:5, Informative)
Re: (Score:2)
Blockchain technology has nothing to do with any of this. It doesn't assist in logistics in any way. It helps untrusted entities prove they submitted data correctly and one time. The end. If you don't trust your logistics workers and truck drivers and warehouse staff, fire them.
So you want them to fire those they don't trust, but don't want them to use a technology which will tell them immediately who they should trust?
Re: (Score:2)
And the blockchain itself cannot be trusted if controlled by a single entity (Nestle), so no help there either.
Nestle doesn't care if Nestle controls the blockchain. Nestle prefers that Nestle controls the blockchain.
Exactly. For PR purposes they might want to make sure it's externally verifiable at some point, but the first step is to get control of your supply chain.
Re: (Score:2)
Re: (Score:3)
If you can't trust your suppliers, then you shouldn't be manufacturing anything. They're not buying their supplies off of Amazon like stupid individuals. They're entering into massive, long-term financial relationships with other companies.
If you aren't auditing your suppliers you shouldn't be manufacturing anything. FTFY
Re: (Score:2)
Remember the melamine being added to animal food, and in milk in asia by chinese companies? Really you can't trust your suppliers even when they're supposedly good in some cases. That should have been an open and shut case for bringing manufacturing back home.
Re: (Score:2)
Re: (Score:2)
That was a good company, and the suppliers were also considered good. Both had a long term history of being good, as in over 20 years.
Re: (Score:2)
Excellent point! The only thing is "Blockchain" (just the term?) sells stock.
If I had a dollar for every company I worked for and put out a press release talking about how they embraced the tech-sector's "Flavor of the Month"? I wouldn't have to work.
Practical? God no. Marketable? Hell yes.
Re: (Score:2)
Or, if you want to prove beyond any doubt that your logs, when audited, have not been tampered. Rather than do a blanket ban of batches 2-27 in case something happened with batches 3, 12, and 21 and you want to be safe, you can just ban those three specific batches because you know with certainty that your logs have never been altered.
Your primary product killing babies is not good for your brand, having absolute certainty in your supply chain seems like a very low-overhead no-brainer and an excelle
Re: (Score:2)
How does it give you certainty? It gives you traceability.
If you find iron filings in bottles from batch F00B4R23 you issue a recall notice that batch. You might test F00B4R20 to F00B4R26 too, because you know they came off the same line. All this is possible now without blockbastardchains, because they do it.
How would a blockchain prevent iron filings getting in there in the first pace?
Re: (Score:2)
I don't think you fully understand blockchain technology. It's not just about BitCoin (although it's a great example of what kind of trust you can place if used correctly).
Blockchain is nothing but a database "log" basically (if you're familiar what SQL logs are) that is cryptographically chained and verified. I agree that in the logistics world, you still have to trust the sensors (whether human or automatic), but if you can verify that your sensors worked and aren't tampered with, then what's in the block
Re: (Score:2)
Re: (Score:2)
I would think it would be better to have some centralized standard number system, like RFID, where the farmers can add to the items they give to the aggrigator who can record and give both to the company making the final
Re: (Score:2)
Re: Here, let me help (Score:2)
Got to disagree with the haters on this one (Score:5, Insightful)
This is one of the few cases where I don't think it's just buzzword marketing. Everyone saying this is an "internal" tracking issue must have never heard the phrase "global supply chain". They're not just trying to track their internal logistics, they're trying to push verifiable tracking out to all their suppliers.
Devil in the details, blah blah ... I know, but this isn't obviously stupid.
Re: (Score:2)
What's a global supply chain got to do with anything?
Say I'm a brewer and I find some of my beer is off. I examine the batches of my product with the problem; I can do this because like any sensible manufacturer of foods & pharma I record the ancestry of my products. I find that they all have one thing in common - hops[1] from supplier X batch Y. That's my internal tracking issue.
I then y
Re: (Score:2)
Say I'm a brewer and I find some of my beer is off. I examine the batches of my product with the problem; I can do this because like any sensible manufacturer of foods & pharma I record the ancestry of my products. I find that they all have one thing in common - hops[1] from supplier X batch Y. That's my internal tracking issue.
Then, because you don't know where they got the bad ingredients, you have to recall everything sourced to them. If they're your largest hops supplier, you might recall a more than half of what's in circulation.
Or, because you've enforced verifiable end-to-end tracking all the way to the end of your supply chain, you know that everything traces back to one farm and you recall 5% of what's in circulation.
Re: (Score:2)
Not my problem. I know who I got them from.
No I don't. Did you see the word batch, I did use it several times.
They might not be for long, if they do it again.
Do you think they keep them all in separate bags or something? In any case, that's the supplier's problem and
Blockchain = profit (Score:1)
"Blockchain! We don't know what it is but every time a company uses it in a press statement their stock price goes up. Let's do it"
This is in response to slavery. (Score:2)
Nestle has used modern slave labor for years. They came up with using block-chain to prove they're coming clean. [theguardian.com] Of course that's just as good as the people making entries on the chain.
They are working to cleanse themselves of the slavery image [nestle.co.uk].
They've used it in their cocoa and coffee supply [awarenessact.com]. They pretty much act like it "slips into the chain" these days, but when they were first found out they seemed more upset they were caught than anything. The older articles are getting harder to find.
Fuck Nestle (Score:5, Interesting)
Trust (Score:1)
I don't trust these new blockchain things. They've not been proven to be safe and I don't want them anywhere near my baby's food. You just lost a customer, Nestlé!
Re: Trust (Score:1)
Oh look, autocorrect put the accent on Nestle.
[Submit]
Oh dear, forgot about Slashdot's automangle.
51%? (Score:2)
Okay, genuine question here: is it not the case that all blockchains are vulnerable to someone simply hiring enough CPU time to outvote the rest of the "miners"? So, anyone wanting to use it to track things like university degrees or food shipments or whatever, has to more or less single-handedly try to keep ahead of anyone who thinks of a way to make money by taking the chain over even for a day using AWS or something. Is that right?
I'm sure this exposes my ignorance but it's been bothering me since someon
Trust and privacy issues will kill this project. (Score:1)