Tag Archives: Provenance

Provenance For Supply Chains

Years ago, I contributed to an earlier effort to create the Open Provenance Model [1], which as since evolved into other efforts, including the larger. Stemming from the challenges of documenting complex scientific processes, data, and conclusions, these rather obscure offshoots of the (itself obscure) Semantic Web have not had the impact I would have hoped.  “Sank without a trace” would not be too far off.

So my eye caught by a new app, called “Provenance”, said to be coming Real Soon Now. And it uses the Bitcoin blockchain to publish provenance.

Provenance-the-app is the brainchild of designer Jessi Baker and others, and seeks to implementing the idea for “supply chains” rather than scientific chains of reasoning. The overall goal is make product sourcing “transparent”, arguing that this will allow people to consume more ethically, and save the planet. This means they need to be a consumer facing app, as well, not just a backend reasoning and query engine.

I’m with you so far. I’m not totally convinced of how well transparency will help save the planet, but it can’t hurt. And Baker is, like me, interested in “making things that fuse digital and physical.”  So let’s have a look.

The Provenance app is highly influenced by blockchain technology, including Ethereum. Their white paper [2] is mainly about in public key signatures, assertions about the perceived merits of “decentralized” systems, and some very detailed ideas about digitally tagging  physical processes and products.

They envision a system of certifications that document the sources and chain of processing for consumer products.  These cryptographically sealed certificates are to be posted on a public blockchain, where they cannot be fiddled with, and where anyone can access them.

The idea is to make it possible to scan a product and quickly retrieve a trustworthy confirmation that it meets your ethical requirements.  Presumably, people will be willing to pay extra for products that have these certificates.

Whether a blockchain per se is critical to this enterprise is arguable. Public key cryptography is certainly useful for establishing chains of trust. But the write-once ledger of the blockchain itself only solves the problem of destroying or hiding information, which isn’t necessarily the biggest problem. There are lots of ways to openly publish data, and, contrary to their assertions, a blockchain isn’t really that much better than other open publications—especially if you have good public key technology.

Provenance-the-app looks to be positioned to make some very important contributions to the overall problem of Provenance, which appears in many guises.  They are working out some hard problems of digital signatures and open data access. This is very impressive.

To the degree that they succeed, their technology might be repurposed into scientific and other record keeping. If you can track a fish from source to table, you can track the “ingredients” of a scientific paper, no?

Looking through the materials available, it appears that they are unaware (or perhaps deliberately reject) they don’t have anything to say about our earlier work and the W3C Semantic Web activity. At least, t They do not explicitly acknowledge it. (See comment below.)

If I may offer advice, I think they would be well served to harmonize with the W3C PROV WG. As “Open Data” folks, they should hew to open metadata standards, no? Second, they might glance at the earlier academic work in which we worked on models for automated reasoning about Provenance. They may not realize it yet, but there are important, difficult problems that they will want to solve in this area. I’m just sayin’.

(Update (12/24/2015):  see comment below, indicating that there is collaboration in progress with W3C PROV.)


 

  1. Luc Moreau, Juliana Freire, Robert E. McGrath, Jim Myers, Joe Futrelle, and Patrick Paulson, The Open Provenance Model. 2007. http://eprints.ecs.soton.ac.uk/14979/1/opm.pdf
  2. Jutta Steiner, Jessi Baker, and Gavin Wood, Blockchain: the solution for transparency in product supply chains. Project Provenance Ltd White Paper, 2015. https://www.provenance.org/whitepaper

Blockchain And Provenance Technology

Reid Williams writes at Conidesk about “How Bitcoin’s Technology Could Make Supply Chains More Transparent”. Specifically, he focuses on the supply chain of food. He describes examples of how this might work, using blockchain, cryptocurrency, and digital signatures.

“At the core of this is the ability to assign identity to people, to organizations, and to goods, to track in a transparent way the provenance of goods as they pass from one organization to the next, and lastly, as goods change hands, to exchange payment between the two organizations.”

His use cases are:

  1. Seeing where your food was grown
  2. Buying local produce in more flexible ways
  3. Seeing how your food got from farm or sea to plate
  4. Rewards programs that cross traditional boundaries
  5. New kinds of markets that create new ways to participate
  6. Knowing exactly what went into your food when it really matters

This all makes sense, even if the examples aren’t earth shattering.

Technologically, he describes a relatively straightforward mash up of mostly old technologies, as he states: identity, provenance, and payment. So what does a blockchain and cryptocurrency add to the picture?

Open identity systems have been under development for years, and digital signatures of various kinds are well developed. These systems do their job with or without a blockchain, and, notably, Bitcoin and other cryptocurrencies are designed to hide identity, which actually defeats the transparency described in the scenarios.

Provenance is very important concept in many contexts. I worked for a while on some pioneering efforts on this topic [1], so I can tell you that Williams is correct about this, and provenance is far more important that just these scenarios. One of the principles of making machine-readable provenance possible is that there must be assertions about “who said so”, as well as who, what, where, when, and how things happened. That is, provenance boils down to trust.

 

In that light, we can see that a blockchain may be useful for publishing key assertions about provenance, but trust in those assertions will depend on other factors, including identities of the alleged parties. Using a blockchain does not actually help with issues about trust, indeed, it is designed to be a so-called “trustless” system.

Generally speaking, provenance systems are designed to reason across multiple trusted sources. I.e., decentralized reasoning about “centralized” authorities. Why? Because expert knowledge is often centralized, and I can define who I trust about what kinds of questions. The local “Better Berry Association” really, really knows about the quality and sourcing of strawberries, so I am willing to trust their ratings a lot more than crowsourced or other ratings..

 

Unpacking this a bit more, provenance systems are about reasoning, taking various digital records (“assertions”), and constructing a trusted story about the object in question. A blockchain is one way to obtain records to reason about, but it doesn’t, in itself, contribute to the reasoning process. A blockchain may be a useful source, of course, if there are no other ways to get the data or the data may be fudged.

 

So, identity and provenance are important, and quite compatible with blockchain technology. But a blockchain is neither necessary nor sufficient for either open identification or provenance. The third technology is, of course, payments, including micropayments, and peer to peer schemes.

 

Here we start to see some new capabilities, which have been universally recognized. We already have co-ops farmers markets and even farm-to-table financial schemes. As Williams discusses, cryptocurrency can simplify and elaborate such schemes, and couple them to other digital technologies such as social media. I don’t think that Bitcoin is necessarily all that different from digital coupons or similar schemes, but there are advantages of scale and universality. The same scheme that I use to purchase local foods will also let me, on occasion, purchase anywhere in the world. (Hey, there ain’t no local grown coffee where I live.)

 

To sum up: Williams gives some believable possibly useful scenarios, and describes a very believable technology. Most of this could be build without a blockchain or cryptocurrency, but it all fits together in a pretty nice package.

 

I have to add a word of caution. A blockchain is only as transparent as the identities of the parties inserting records (“assertions”) on it, and the information on the blockchain is only as trustworthy as the parties involved.

If you want to see what I’m talking about, glance at the Coindesk report by Stan Higgens about “Former Exec Hits Back at OKCoin Amid Contract Dispute”. Despite the use of digital signatures and “transparent” contracts, there is still a savage dispute, little different from so many other such crack ups.  Using a blockchian or Bitcoin was certainly not sufficient to prevent this dispute.

The blockchain is neither necessary nor sufficient for creating trustworthy systems, but it may be a very useful tool—along with other technologies.


 

 

 

 

 

 

 

  1. Moreau, Luc, Juliana Freire, Joe Futrelle, Robert McGrath, Jim Myers, and Patrick Paulson, The Open Provenance Model: An Overview, in Provenance and Annotation of Data and Processes. Springer, Berlin, 2008, 323-326. http://dx.doi.org/10.1007/978-3-540-89965-5_31

 

Cryptocurrency Thursday