Blockchain Can Tell You where that Raspberry really came from.
But lack of data, power consumption issues and the cost to farmers point to a Technology that may not be ready.
I came across a fascinating story about a raspberry recently.
Well, not one raspberry, but a whole shipment of them.
Dave Sherwood from Reuters wrote a illuminating article about how raspberries were involved in a case of food fraud.
You can read all about it in Mr. Sherwood’s piece, but in sum, in January of 2017, Chilean customs officials received a tip from a whistleblower who informed them that Chile’s crop of raspberries — fruit that is grown during the winter months in the South — was under threat. When the officials raided the offices of Frutti de Bosco, they discovered an operation designed to pawn off Chinese raspberries as Chilean in an operation that spanned three continents.
Raspberries grown in China, then flash frozen, were shipped to New Zealand via a middleman. Frutti de Bosco in Chile bought the berries from the middleman and had them shipped to Chile. There were then moved to a packaging plant, re-packaged and labeled as being both ‘premium and organic’ and as originating in Chile. The berries were then shipped again to the Canadian market, specifically to food wholesaler based in my home town, Montreal, for retail sale.
I’m a sucker for field berries. I know that berries of all sorts are shipped from Chile and Peru during the winter months, with their final destination being berry-starved consumers in cold Canadian cities. I may have even bought some of those ‘Chilean’ raspberries.
But the only way I know where the berries I buy come from is the word of the person selling them to me, and the signage the vendor uses. I have no way to actually validate where those berries came from through other means.
Reading the Reuters article made me realize the insidiousness of food fraud and how the provenance of the food we eat is difficult or impossible to validate as a consumer, but also, how little attention it gets from Canadians. Apart from media coverage for food recalls, food fraud gets little coverage. The Agri-Food Analytics Lab at Dalhousie University published a recent survey where 74.9% of Canadians generally feel that food products in Canada are safe.
And yet, some of those ‘Chilean’ raspberries made Canadians sick, leading the Canadian Food Inspection Agency (CFIA) to issue a recall on those berries coming from China. What they didn’t realize until later was that potentially lethal berries were coming into Canada via the Chilean backdoor.
Promoters of Blockchain technology have been quick to inform the food world that they have a solution for this problem. Blockchain applied to the food supply chain will lead to full transparency in food provenance, they say, not only for the food wholesaler and retailer, but for the consumer as well and ensure that food imported into Canada is safe for consumption.
As a consumer, and as someone working the tech industry, the promises of blockchain enthusiasts makes me ask — really? How does that work?
First, the basics about blockchain.
Blockchain is a system for recording ledgers, or records, of transactions in a de-centralized and encrypted manner. The records can be anything — financial records, certifications, emails, records of product shipments, accreditations, etc.
Rather than have such transactions reside in a centralized database controlled by a centralized authority, the information is instead distributed in a decentralized manner over thousands of sites, called blocks.
Each block contains a certain amount of information or transactions, and is linked to the block before it. As each block of information is linked, forming a chain, the term blockchain is used to describe the entire sequence of transactional information for a particular industry or application.
The key challenge with a distributed manner of recording information is how to permit multiple parties to create, verify and update information within the many blocks of information within the chain.
At least three basic elements are required for blockchain to function:
- A peer-to-peer computer network that is open to everyone, and accessible by everyone. We have this today in the form of the Internet. It allows contributors to share information on the same blockchain from wherever they are located on the globe.
- Cryptography is required to ensure that contributions to the blocks are unaltered, and verifiable as not coming from bad actors. This is done through the use of a cryptographic function that translates every input into a fixed-length string of numbers. Each block in the chain is also identified by using a unique numerical sequence, and each block also contains the identifier of the previous block. This way of identifying blocks makes the data within the blocks immutable, or impossible to change, as to change one block means changing all the previous blocks.
- A Consensus algorithm, or rule, is required so that everyone agrees on how to open the ‘gate’ to access the chain to add information to blocks, and how information in the blocks are verified. In Bitcoin’s case, they use a Proof of Work algorithm, whereby persons wishing to add or change information have to solve a mathematical problem, which requires data mining and a lot of computational power to solve (I’ll come back to this particular point below).
However, blockchain technology should only be used if there is a demonstrated case to be made for a decentralized, distributed network. If there are no advantages to decentralization, there’s no use in implementing blockchain technology; in that case, a centralized solution works probably more efficiently.
Before a particular sector implements blockchain technology (like the food sector, for this article), some obvious questions need to be asked:
- Is there a need for a shared common database?
- Are there multiple parties involved (usually from different entities)?
- Do the parties involved have conflicting incentives and/or are not trusted?
- Are the rules governing participants uniform?
- Is there a need for an objective immutable log?
- Do the rules governing transactions change infrequently?
With the story of our ‘Chilean’ raspberries in mind, we can provide answers to the questions above fairly easily:
- There is a need for a shared, common database with food supply chains. A distributed database would permit all members of the supply chain to access the database to input their particular part of the data from harvest to a retailer at the final destination, making tracing the provenance of a particular food much easier, but also, more trustworthy, given that records and certificates can be verified by purchasers and government regulators;
- There are many parties involved in the transshipment of foods within a global supply chain, even within a very local food chain with a smaller jurisdiction like a Province. From the farmer to the packer to the shipper to the wholesaler to the retailer — the chain is pretty much the same whether between a farm in rural Québec and a vendor in Montreal, and a farmer in rural Chile and that same vendor in Montreal;
- I wouldn’t go so far as to say that those multiple parties have conflicting or untrustworthy incentives, but they do have different interests and the current systems are vulnerable to attack by bad actors, like the company in Chile. A blockchain system of maintaining and updating transactions might help smooth out the differences in a more open manner between them all and prevent bad actors from hopping onto the supply chain to disrupt shipments, or gain an advantage;
- The rules governing the different parties of a food chain are not necessarily uniform today, but, there are certain norms that can be applied to a food chain, such as GS1 standards, which have been adopted by more than a million organizations worldwide, and can help all members ‘speak the same language’ with regards to what products are being shipped and how they can be tracked. Such adoption would go a long way in unifying the information used by all parties within a food supply chain;
- Given the ‘Chilean’ raspberry example — and many hundreds of other examples of food fraud — there is a definite need for an immutable, objective log of transactions along the entire food supply chain to protect the health of consumers, and assist consumers is learning more about where their food is coming from. Allowing CFIA inspectors in Canada to validate the organic certification of the raspberries would’ve been enough to stop their distribution to retail shops;
- Last, rules governing transactions will change over time; trade agreements, changing environmental conditions, changing political considerations, changing habits of consumption all drive change in how the transactions are done. But, for the most part, the logistics in getting an item from a farm to a retailer are pretty consistent and certification details are being agreed to in international fora for application around the globe.
Based on this short analysis, it seems that blockchain — as applied to a food supply chain — is suitable for this sector, and would drive significant benefits in terms of traceability and transparency.
Blockchain is not omnipresent and cannot magically watch a particular product from the farm to the plate. Like any data storage system, blockchain needs inputs. It needs humans to interact with the platform. And this is, in my mind, where the applicability of blockchain gets less obvious.
There are limits to what farmers can actually input into a blockchain system today. The information inputs around products produced by smaller farmers, in particular, are not standardized. Farmers lack the tools and technology to gather the data that might be useful for a blockchain system. Farmers grow natural products; their consistency, makeup and content change with each crop and is highly dependent on the inputs (or, in the case of a farm practicing regenerative methods, the lack of artificial inputs) that are used to produce a particular crop.
Blockchain can’t verify that an egg operation is truly cage-free or what that cage-free operation really looks like. But it might offer a way for farmers to get more information to consumers. Farmers who don’t sell their food to a farmer’s market or get an opportunity to interact with consumers, often struggle with how to engage with the public, looking for ways to explain how and why they grow food the way they do. Blockchain enables farmers to get data to consumers, but what is needed more from a consumer level is the context in which that product was produced in the first place. That’s what consumers really need to make informed decisions about their food.
Take the recent Buttergate issue. If a consumer purchasing butter would be able to know at the point of sale which farm their butter came from, the feed used by the farmer to produce the milk at any particular time and how that farmer treats her animals, that would be a powerful tool for consumers to influence their purchasing decisions but also, a powerful tool for the farmer to be further incented to sell the best product they could. Dairy farmers are already incented to produce a quality product, but they lack the means to really communicate their essential context directly to the consumer. Much of their product is purchased by large, consolidated producers, who use the raw product of milk to product any number of other derivative products, such as cheese and yogurt.
A blockchain system that focused on dairy production might be the answer. Or one focused on the importation of fruits and vegetables or the production of fruits and vegetables more locally. And another on the production of beef. And so on.
There is another issue that needs to mentioned — power consumption of blockchains.
There’s been plenty written recently about how much electricity Bitcoin requires.
There is a good paper written by Sedlmeir et al, in June, 2020 that provides detailed explanations of why this is so. Here is one paragraph from that paper that sums up well why Bitcoin is so power-hungry:
Bitcoin’s key innovation was to provide a suitable consensus mechanism for the use in this scenario. Specifically, Bitcoin combined several well-known concepts from cryptography to form the so-called PoW [Proof of Work]. This refers to the right to create a new block from a subset of queued transactions when one finds a solution to a cryptographic, computationally intensive puzzle. The process of searching for a solution is called “mining”. This results in coupling the voting weight to a scarce resource — computing power and thus energy — and hence prevents Sybil attacks. The mining process is economically incentivized in that participants are rewarded for every valid block that is found and disseminated. The reward typically consists of a certain amount of the associated cryptocurrency and the fees for the associated transactions. The value of the former is proportional to the cryptocurrency’s market price, so the success of cryptocurrencies on financial markets in the last years has provided a very strong incentive to participate in mining. In turn, this has led to an enormous energy consumption associated with the underlying PoW blockchains.
Based on calculations by Sedlmeir et al, the average electricity consumption by the top five Proof of Worth cryptocurrencies is around 110 TWh per year. Since I wrote this, it has risen to 120 TWh per year, according to Cambridge Bitcoin Electricity Index. As a comparison, Hydro Québec sold 171.4 TWh of electricity into the Québec market during all of 2020. To put that consumption into perspective, a single bitcoin transaction uses the same amount of electricity that an average American household consumes in one month. Almost 70% of the world’s data mining activity is occurring in China, which will certainly have an impact on that country’s ability to curb its greenhouse gas emissions over the next decade. It’s getting to the point now that cyptrocurrency firms are building their own power plants, to guarantee the delivery of electricity to support their mining operations.
And as Bitcoin — and other cryptocurrencies Proof of Worth calculations continue to grow — concerns are being expressed on the very sustainability of blockchain, given how much electricity it is diverting from electrical grids.
There are alternative methods of validating/adding to the information in blocks:
- Proof of Stake — rather than focusing on a scarcity of computational resources, PoS relies on the scarce resource of capital. There is a random mechanism that determines who is allowed to build (“mint”, “forge”, “bake”) and attach the next block. With the help of this mechanism, the probability of being selected is linked to the amount of cryptocurrency that the node has deposited and locked (“staked”) for this purpose. The deposit also incentivizes the node to stick to the rules of the network, as any misbehavior detected will lead to the node losing this deposit. The advantage of PoS is that it does not involve any computationally intensive steps such as solving the cryptographic puzzles in PoW.
- Proof of Identity or Proof of Authority — this is applicable to what’s called ‘permissioned’ blockchains, where the users are authenticated prior to being permitted to use/add to the blockchain in question. Each user has one vote, they are all known, and they gain entry through a variety of mechanisms, where the more secure they are, requiring more and more computational power.
Sedlmeir et al did not go into detail on assessing the power consumption levels of these other algorithmic methods of accessing blocks, and information is hard to find on just this element.
Regardless of the method used to access blocks, it is clear that power consumption and the footprint from an environmental point of view needs to be taken into account as part of any assessment to deploy blockchain.
Where does this leave my basket of imported raspberries?
Well, still subject to current levels of authentication — inspection by Canadian authorities, diligence on the part of importers and following the news on any health issues that may be tied to the consumption of tainted berries.
This is still a new field. But there are companies in the agricultural domain using blockchain to implement traceability and transparency initiatives:
- Ripe.io — Ripe.io has conducted small pilots that have showcased the end-to-end chain from seed to mouth of fresh produce, capturing data related to field growth conditions, temperature data within distribution trucks and information extracted from sensors on companies’ food processing machinery. Ripe.io can then build a model based on the data to incorporate smart contracts around the process.
- Provenance (UK) — is actually employing blockchain for over 200 food and drink producers in Europe. For example, Napolina tells the story of how and where they source their tomatoes for their canned products, including the labour practices of their tomato farmers; the Princes Group provides information to the consumer about the sourcing, treatment and fishing methods used in the marine products it sells; and the International Pole and Line Foundation implemented a solution for providing digital proof of sustainability and slavery-free marine products for export to the UK, European and US markets.
Large retailers in the United States, such as Walmart, are implementing blockchain solutions for their leafy greens, forcing their suppliers to jump on board and adopt the technology.
It would seem to me that the CFIA should be working harder at encouraging Canadian companies to implement such solutions to make their work less relevant. If more of the food industry were to use blockchain solutions, incidents of food fraud and questionable and dangerous food quality might decline, making the work of the CFIA inspectors that much easier. The CFIA certainly has this on their radar, and has conducted feasibility studies on the use of blockchain in the livestock industry to improve traceability and reporting.
More than that, however, food producers themselves ought to be looking into the use of blockchain as a way of supporting their brand and improving the quality of information that consumers can access on the products they buy.
Based on everything that I’ve seen, there are at least three issues that are potentially holding up wider implementation of blockchain in the food production industry in Canada:
- Lack of data. Farms, in particular smaller farms below a couple of hectares in size, simply don’t collect data on every aspect of their operation. Even larger farms lack the tools necessary to gather key information indicators at the farm level that might be useful in any kind of blockchain solution. Will 5G as applied to the farm help in this way? I’m not sold, as I indicated in another article on this topic.
- Electricity consumption. Many of the current methods used to validate the security of the individual blocks in the chain are driving up the consumption of electricity to unsustainable levels. In addition to developmental work on different types of authentication algorithms, some companies are even trying to re-architect blockchain, such as Red Belly Blockchain. Current PoW algorithm-based blockchain systems are designed for all data miners to validate all transactions in the chain, thereby driving up power consumption considerably. Red Belly architects the blockchain differently, by separating the validation process so that, one CPU validates block A, while another validates block B, and so on, considerably reducing the computational effort required to validate transactions.
- Cost. For farmers to be a part of a blockchain system, and effort required for them to input the data, the cost today is considerable. They would have to re-design their operations to be better at capturing specific types of data that can be entered into a particular blockchain for their industry. Given the margins that farmers operate under, this is not likely to happen without some assistance; for smaller farmers, this may be out of reach completely. For food importers and wholesalers, this ought to be a requirement, frankly. If other individual producers can do this around the world, then there is no reason our own cannot as well.
I suspect that it may be a while before I can ask my vendor at the market for the QR code so I can scan it and check quickly the provenance of the raspberries she is selling me in the middle of winter. Or, if I can scan a similar code on my litre of milk and get similar info regarding the farm from where it came, and an overview of the animal welfare practices of that farmer.
And frankly, I’m not sure that I necessarily want to participate in a solution that is potentially diverting a large amount of electricity from people who need it for basic needs, such as heat, light and food preparation.
Blockchain proponents have to be more transparent in explaining to consumers how ethical their solutions are, so that wider acceptance and adoption can occur. After all, everyone wants to buy food that is good for us, is produced using sustainable and ethical practices and is not going to kill us if we eat it.
But the solution can’t be worse than the problem. If it is, then we’ll be imposing more of an environmental burden on society as a whole, and it won’t be sustainable in the long run.