Fairness, Transparency and Traceability in the Coffee Value Chain through Blockchain Innovation

Abstract

Coffee is both one of the most widely consumed beverages in the world and one of the most important internationally traded commodities. However, the coffee value chain is opaque and unbalanced: less than 10% of the $200 billion industry value remains in producing countries. First, this article introduces the idea of a Commodity Fairness Index used to measure the inequality, or economic imbalance in a commodity value chain, and calculates it in the case of Colombian coffee. Secondly, the paper describes the system architecture of a web app built upon Hyperledger Fabric that paves the way to improve coffee farmers' lives by bringing transparency and traceability into the entire value chain and improving its fairness as a result. The application of blockchain technology to the coffee industry enables inclusive business models that reward quality and hard work, and in turn translates into greater trust, confidence and fairness across the entire industry as well as among end consumers.

Full text

Fairness, Transparency and Traceability
in the Coffee Value Chain through
Blockchain Innovation
Federico Miatton∗†, Laura Amado†
∗Universitat Pompeu Fabra, Barcelona, Spain, †Fantine.io, Bogot´
a, Colombia
Abstract
Coffee is both one of the most widely consumed beverages in the world and one of the most
important internationally traded commodities. However, the coffee value chain is opaque and unbalanced:
less than 10% of the $200 billion industry value remains in producing countries. First, this article
introduces the idea of a Commodity Fairness Index used to measure the inequality, or economic
imbalance in a commodity value chain, and calculates it in the case of Colombian coffee. Secondly, the
paper describes the system architecture of a web app built upon Hyperledger Fabric that paves the way
to improve coffee farmers’ lives by bringing transparency and traceability into the entire value chain
and improving its fairness as a result. The application of blockchain technology to the coffee industry
enables inclusive business models that reward quality and hard work, and in turn translates into greater
trust, confidence and fairness across the entire industry as well as among end consumers.
Index Terms
Coffee Value Chain, Commodity Fairness Index, Blockchain, Distributed Ledger Technologies, DLT,
Supply Chain, Colombia, Technology for Social Good, Inclusive Business Models, Hyperledger.
Proceedings of the IEEE 2020 International Conference on Technology and Entrepreneurship – Virtual (ICTE-V), Apr. 2020.
Correspondence: rand @ fantine.io.
https://doi.org/10.1109/ICTE-V50708.2020.9113785
1

Mexico
Guatemala
Honduras Colombia
Peru
Brazil
Vietnam
Indonesia
Ethiopia
Uganda
India
Nicaragua
China
Costa Rica
Kenya
Papúa new Guinea
Tanzania
El Salvador
Ecuador
Laos
Madagascar
Rwanda
Panama
Fig. 1: The yellow “bean belt” represents the region of the world where coffee is produced; the
blue band highlights the main consuming countries.
I. INTRODUCTION
More than 2 billion cups of coffee are consumed around the world everyday; coffee is one
of the most important commodities traded worldwide and the coffee industry generates around
$200 billions yearly [1]. Coffee is one of the most widely consumed beverages in the world,
and not only plays an important part in the daily routines of a significant share of the world
population, but it also has a significant social and economic impact for the families that produce
it, who are mostly smallholder farmers running plantations of less than 5 ha. Due to agroclimatic
conditions, 90% of the world’s coffee production takes place in 45 developing countries where
it employs 25 million farmers and provides a livelihood for more than 100 million people [1].
The so called “bean belt” is highlighted in yellow in Figure 1: this is the band around the
middle/southern region of the world, roughly bounded by the Tropics of Capricorn and Cancer,
that is ideal for coffee production. Over the past few decades, unfortunately, coffee has not
been a stable and profitable source of income for farmers. In fact, even though there are widely
varying levels of quality among beans, coffee is primarily bought and sold as a commodity
crop– meaning its sale value is tied to international futures markets, such as the New York
Stock Exchange, and is not reflective of an individual farmer’s true costs of production.
2

Low and volatile prices of green beans are not the only challenges that small growers face.
Other economical issues include limited market access for producers, lack of product and market
information, and exchange rate volatility [2]. Ironically, over the past two decades, the price paid
by consumers for a cup of cappuccino has grown 150% while the price paid to farmers for one
pound of beans has stayed flat or even decreased over the same period.
Indeed, in the last decade, coffee has been the only commodity that has been depreciating. As
an example, in March 2020 coffee was exchanged at $1.03 per pound, the same price marked in
February 1976 (not adjusted for inflation) [3]. The result is that coffee producers are continuously
living on a loss and depending on aid from governments. They are notoriously poorly paid in
a relatively rich industry where just between 5%-10% of the global industry value remains in
producing countries [1]. The reasons are manifold and complex [2]. Furthermore, the common
feeling that certifications, such as Fair Trade, can solve the problem is actually a misconception,
because certifications are paid for by the farmers and don’t reach the majority of growers. So,
why instead of thinking about social responsibility practices and certifications, which in most
cases don’t reach the majority of growers, don’t we redesign the entire coffee value chain to
make it completely transparent, efficient and able to generate win-win economic transactions?
[4] This question summarizes the motivation for this article.
The rest of the paper is organized as follows: Section II presents some related work; Section
III details the coffee value chain and introduces a novel Commodity Fairness Index that can be
used to measure the inequality among players in a commodity value chain and calculates it in
the case of Colombian coffee; Section IV introduces the main problems the coffee industry is
facing, discusses the design requirements and the system architecture of our solution, and finally
Section V draws the conclusions.
II. RELATED WORK
In this section, we present some of the related work found in the literature on the coffee value
chain and on blockchain applications for food and agricultural supply chains. The application
of blockchain to supply chain management has been a topic of much research in the past years,
and promises to deliver up to 15-20% cost savings along the supply chain. Even the World
Economic Forum has recently published a series of white papers covering specific governance
considerations for decision-makers deploying blockchain solutions throughout international trade
and supply-chain systems [5].
3

The authors in [6] present AgriBlockIoT, a blockchain based traceability solution that integrates
data from IoT devices along the value chain. They define a use case to track crops from farm
to table and compare the performance of two implementations in Ethereum and Hyperledger
Sawtooth. Another proposal [7] makes use of smart contracts on Ethereum to execute and
keep tracking of transactions. Similarly, another product traceability system, also based on the
Ethereum blockchain, is proposed in [8], in which all product interactions are recorded and kept
in the ledger through the use of smart contracts, thereby creating a sequence of events that
represents the entire product history and which can trace it back to the source.
An example of commercial application is given by [9], in which blockchain technology and
sensors are used to track caught fish. The objective is to ensure conformity to standards at
origin and throughout the chain, which is implemented in a secure way on the blockchain
without relying on a centralized infotracing system. Relying on blockchain permits to solve
the problem of the double-spending of certificates without a central verification authority. By
polling the unique ID of the object enables anyone to visualize the details about that particular
item. Another business application is provided by the IBM Food Trust [10], a network that
leverages Hyperledger Fabric [11] to connect participants across the food supply chain through
a permissioned, permanent and shared record of food system data. The result is a solution that
promises to increase food safety and freshness as well as to minimize waste.
Finally, a primer on the global coffee chain is provided by [12], which analyses in detail the
coffee value chain in the past forty years using methodological instruments drawn from the global
commodity chain analysis developed in the context of political economy. The article describes
the changes that affected the global coffee chain dramatically, as a result of deregulation, new
consumption patterns, and evolving corporate strategies. It also examines the various shifts that
occurred in coffee trade “regimes” throughout the past decades and focuses on market power
and corporate strategies in the current configuration of the global coffee value chain.
III. GLO BAL COFF EE VALUE CHAIN & DE FIN ITION O F THE COMMODITY FAIRNESS INDEX
The coffee value chain is extremely complex, opaque and it involves a lot of hands. Figure 2
shows a view of the value chain, depicting all the main actors involved between the production
and the consumption of coffee. One additional step that is not shown in the figure is the potential
storage and warehousing of beans, which may occur at almost any stage of the chain, and may
either involve third parties or may be taken care of by the same players shown in the chain.
4

Producers
Mill Transport
Commodity
Traders Exporters
Importers
Shippers
Brokers Roasters
Packagers Distributors
End consumers
Fig. 2: The Global Coffee Value Chain.
Like in many commodities, the coffee value chain too starts with a large number of growers
that produce widely different quantities and qualities of coffee. In the case of coffee, there are
around 25 million farmers globally [1], of which around 500’000 in Colombia alone, mostly
smallholders. The beans reach global markets through a series of activities and processes that
make them suitable for incorporation into industrial operations [1], including processing at the
mill, packaging into coffee bags suitable for export, and transportation. In the following, a brief
description of each actor is provided.
1) Producers. Grow and process coffee, often with the use of third parties that do the fieldwork
and help them with all matters in the farm. Growers have the possibility to sell their crops
in cherries, wet coffee, or in parchment. Producers can sell coffee directly, to local traders,
to exporters at farm gate, to (or through) cooperatives.
2) Mill. The processing mill is where the parchment is removed from the bean. A coffee bean
has 5 layers of skins of different thickness and parchment is the last one. The process of
removing the parchment skin from the beans is called coffee hulling and the result is called
green (or raw) coffee, which is packaged into bags ready to be roasted, sold or shipped.
3) Brokers/Commodity Traders. Trade coffee through a variety of financial instruments, mostly
futures, impacting the price of the crop on international markets. Up to 40% of all coffee
trades proceed from financial transactions.
4) Transporters/Shippers. Deal with the transportation of the crops; physically move the beans
on the road, by plane, or by sea.
5) Exporters. Buy coffee directly from farmers, from cooperatives or from local traders. If
necessary, exporters may also take care of the hulling coffee process according to the
specifications provided by the importers. They also take care of the domestic logistics,
financing, and can sell it Freight On Board (FOB) at the port of origin.
5

6) Importers. Trade green coffee beans and bring them to the market for blends or single-
origin, depending on the requirements of roasters. Importers may also deal with the
financing of the crops, logistics and procurement at origin.
7) Roasters. Buy green (i.e. unroasted) coffee beans and roast them according to the tasting
profile that appeals end consumers. The type of roasting applied to the coffee (e.g. light,
medium, dark) strongly depends on the local market where the coffee is to be sold.
8) Packagers/Distributors/Retailers. Buy roasted beans or grinded coffee from roasters, store
it and have it ready for selling, normally in national markets.
A. The Commodity Fairness Index
This section introduces a Commodity Fairness Index (CFI) that provides a quantitative measure
of the equality (or fairness) among the main players in a commodity value chain. The motivation
comes from the Gini coefficient [13], which provides a statistical measure that is used to evaluate
the distribution of the income among the population of a country. Hence, the Gini index helps
measuring the inequality of the income of a country’s population. In a similar way to the Gini
index with income inequality, it is possible to measure the fairness in a commodity value chain.
The CFI tries to evaluate quantitatively the imbalance of a commodity value chain. In the
specific, the metric that we want to evaluate is the value capturing element of the players in a
given commodity value chain. Value capture is one of the three essential elements of a well-
defined business model [14]. Hence, the question that the CFI is trying to answer quantitatively
is: How unbalanced is the coffee value chain from the perspective of capturing value? In reality
one may also evaluate a number of additional metrics that characterize a commodity value chain.
Because of this, we should perhaps speak about a Commodity Fairness Indexes Set, rather than
just one single index. Nonetheless, we start by introducing the CFI, and we leave to future
work the measurement of additional metrics that we believe could aid in further evaluating a
commodity value chain.
The starting point for the definition of the CFI are Lorenz curves adjusted for this objective.
The original Lorenz curves used to calculate the Gini index plot the share of population against
the share of income received [13]. Here, instead, we are interested in plotting the share of the
population vs. the share of the value captured in the value chain. In the context of a value chain
the notion of population refers to the set of actors that have a stake in the value chain. In our
definition, the single atomic element for calculating the population is given by an individual or
6

100
% of players in the value chain
% of value captured
100
|
|
A
B
0
Commodity Fairness Index
Fig. 3: The Commodity Fairness Index is defined as the ratio between the areas of regions A
and A+B, as labelled in figure.
company that acts in her economic interest as an independent player in the value chain, interacting
with others. Atomic elements are, for instance, a single farm selling coffee, an exporting or
importing company (even though several people may be working for the same firm), a roastery.
The Commodity Fairness Index is defined in Equation 1 as the ratio between the area regions
of Aand (A+B)as illustrated and labelled in Figure 3.
CF I =A
A+B(1)
In order to calculate Aand B, and plot the Lorenz curves of Figure 3, the percentage of value
captured per each player in the value chain is calculated over the total economic value generated
in the value chain. Clearly, the share of value captured in the y-axis must be evaluated in the
same currency for all players to ensure consistency. One additional refinement of the index may
come from calculating Aand Bat purchase parity power and we leave this to future work. The
CFI coefficient measures the inequality of the value capture part of a commodity value chain. A
CFI coefficient of zero expresses perfect equality, where all parties capture the same percentage
of value. On the other hand, a coefficient of one expresses maximal inequality, where only one
player captures nearly 100% of the value and the others none.
7

B. The Colombian Coffee Fairness Index
In this section we evaluate the CFI in the case of Colombian coffee. The data used to calculate
the CFI in this case are shown in Table I; these data come from various sources, mainly [2],
[12], [15] as well as from additional analysis we did on field in Colombia as well as our own
market research. The share of value captured by some players, especially those in consuming
countries, namely importers and roasters, may vary a lot and strongly depends on the consuming
market. The figures shown in the table refer to the European market and are averaged. Figure 4
shows the Coffee Fairness Index in the case of Colombian coffee, which turns out to be:
CF I = 0.8588 (2)
Colombian Coffee Value Chain
Player in the Value Chain Share of Value Captured Share of Value Chain Population
Producers 5% 89%
Mills/Processors 2% 5%
Exporters 9% 1%
Transporters/Shippers 7% 1%
Importers 32% 1%
Roasters 45% 3%
TABLE I: Data used to calculate the Colombian Coffee Fairness Index. Sources: [2], [12], [15],
own analysis and additional data coming from market research.
The CFI index given in Equation 2 is much closer to one than to zero, showing that a value
chain described by the data of Table 1 can be considered unbalanced and unfair. According to the
table above, in fact, almost 90% of the population is capturing barely 5% of the value created.
Because of this, one might have expected a similar result, at least qualitatively. Nonetheless, the
CFI helps in providing a quantifiable measure of the balance and fairness of the value chain.
Actually, Equation 2 simply highlights a situation that is well known in the coffee industry. In
fact, recent estimates show that at least 44% of the world’s smallholder coffee farmers are living
in poverty [16] and the reasons are manifold. Apart from the unequal income distribution, coffee
growers are bearing most of the industry risks. In addition to the price volatility and exchange
rate risk, producers must also shoulder the risk of losing their crops due to bad weather and to
fluctuating costs of production [17].
8

0.0 0.2 0.4 0.6 0.8 1.0
share of value chain population
0.0
0.2
0.4
0.6
0.8
1.0
share of value captured
CFI: 0.8588
Fig. 4: The Colombian Coffee Fairness Index.
Furthermore, in Colombia specifically, farmers experienced an alarmingly reduction in land
ownership, which occurred throughout the past fifty years [15]. In fact, in 1970 53% of farms
were smaller than 5ha, figure that grew to 79.7% in 1997. More recently, namely in 2018,
this percentage further grew to 94%, of which 90% accounted for farms smaller than 3ha. One
additional metric that could aid in measuring the fairness of a commodity value chain could be
provided by the amount of risk shouldered by each player in the value chain versus the value
received from taking that risk. We leave this to future work.
IV. TOWARDS BRINGING FAIRNESS AND TRANSPARE NCY INTO TH E COFF EE INDUS TRY
As we have seen in previous sections, producers are the weakest link in the coffee value chain.
While they shoulder most of the industry risks, they capture the least value from their economic
activity. However, coffee is a business for anyone involved in its value chain; it is a business for
farmers as it is for roasters, importers and exporters. The goal of farmers is a reasonable profit,
not the aid, not the charity, not to be at the center of social programs [4]. In the end, farmers are
just another value-chain business partner and should be treated and rewarded accordingly [17].
We believe that the entire coffee value chain needs to be reinvented to become more efficient
and able to generate win-win economic transactions even for the weakest players in the chain,
the producers.
9

This section illustrates how the application of blockchain technology to the industry can
overcome some of the challenges described previously and provide a way to tackle the im-
balance in the coffee value chain highlighting a path that leads to creating a more balanced
and sustainable industry while simultaneously creating additional value in both producing and
consuming countries as well as at the consumer end. The first step towards doing so is to bring
transparency into a complex, opaque and closed value chain. Blockchain, or distributed ledger
technology (DLT), is a distributed environment that combines different well-known technologies
(such as hash functions, asymmetric cryptography, digital signatures, peer-to-peer networking)
to store any type of data in a distributed fashion among its participants. The data is organized in
transactions that are embedded into blocks. The DLT encodes the rules that nodes must obey in
order to participate in the network and create transactions that can be accepted into the blockchain
ledger. The ledger acts as a shared record of trust agreed upon by all participants.
A blockchain environment is called public if anyone can freely join its network as a node.
Otherwise, if only authenticated nodes are allowed to become part of the network, then the
blockchain is called permissioned. While public blockchains are still important and valuable
options for certain use-cases, we believe that permissioned DLT systems are in general more
suited to business applications, especially where some level of confidentiality or secrecy of
the transactions –or the transactions’ details– occurring in the system, must be ensured among
participants.
A. Design Requirements or: Why Blockchain?
We argue that the first step towards improving the balance and the fairness of the coffee
industry is to bring transparency into a complex, opaque and closed value chain. A permissioned
blockchain platform provides the ideal platform to achieve that. This choice follows from the
design requirements that were defined for the platform that we are building, and which are briefly
summarized in the following.
1) Delivering full transparency into the coffee value chain in order to enable everyone to
know exactly where the coffee came from and the conditions under which it was shipped.
Enabling full price transparency to ensure producers are paid fair prices for their crops.
2) Enabling process optimization and collaborative demand management between farmers and
buyers, which will enable to build better forecasting models, identify bottlenecks and in
turn unlock supply chain efficiencies.
10

3) Delivering end-to-end traceability and verified provenance throughout the entire supply
chain, in order to certify the coffee’s origin as well as its quality and freshness.
4) Enabling farmers’ visibility and making inclusive business models possible. This will in
turn create a positive virtuous cycle in which growers that produce great coffee are rewarded
accordingly, further motivating them to produce even better coffee.
5) Empowering producers to securely share audits and certificates, in order to prove that they
utilize and promote sustainable practices in their farm.
6) Digitizing the complex and time-consuming paper trail required to keep and facilitate as
the goods move along the supply chain from farm to the final consumer.
7) Ensuring data confidentiality, avoiding that anyone in the chain accesses business infor-
mation that should be kept private.
Changing the status quo in order to redesign the entire value chain to make it more transparent,
efficient and more inclusive is a collaborative process that requires the effort of all the players
involved in the supply chain and this won’t happen overnight. However, we believe that this is
a necessary step in order to ensure the sustainability of the coffee industry in the long term.
Enabling end-to-end traceability and full transparency into the whole coffee supply chain is key
for a number of reasons.
First, end-to-end traceability guarantees provenance and enables users to gain a clear view
on how long the coffee has been travelling. In the case of coffee, we tend to think that freshly
roasted coffee is important; however, roasting a freshly harvested coffee is even more important.
Secondly, being fully transparent about the quality and origin of the coffee enables to build trust
into the coffee chain, and gain consumer confidence. Furthermore, full transparency is essential
to ensure that growers are paid fairly for their crops and helps to ensure conformity to standards
at origin and throughout the chain and certify sustainability practices along the supply chain.
At the same time, full transparency also permits to discover inefficiencies and lack of sustain-
ability across the supply chain, and can also prove what players in the chain utilize and promote
sustainable practices. This is critical in an era when end consumers are becoming more and more
“sustainable-conscious” and are demanding that their retailers become more transparent about
their sustainability efforts [18]. Additionally, full transparency translates into full accountability
for every player of the chain, reducing waste and mitigating the possibility of food fraud. Finally,
the combination of verified provenance with complete end-to-end traceability and transparency
enables inclusive business models, in which coffee growers can be rewarded for the quality of
11

Broadcasting
Company 1 CA
Distributed
memory-key
Database
Company 1
Database
Database
Rest API
Company 2 CA
CA Data
Storage
Queue
Manager
Company 2
Ordering service
Node 1 Node 2
Broadcasting
Hyperledger
Fabric Blokchain
Fig. 5: System Architecture of a web app built on top of Hyperledger Fabric.
their work, enabling them to break free from the speculative pricing of commodity markets, on
which they have otherwise no control.
B. System Architecture and Message Flow
In this section we describe the system architecture of a web app built upon the Hyperledger
Fabric blockchain framework [11] that enables transparency and traceability into the entire coffee
value chain. Figure 5 depicts the simplified system architecture and the message flow of the
defined architecture. A brief description of the message flow in the system is provided in the
following.
1) The message flow starts with a user interacting with the web-based app either through
her desktop or through her smartphone. The request to fetch the data –or to write the
data if she is allowed to do so– is sent to REST API, which is responsible for creating
a communication link between the server and the web/mobile based client. The API
acknowledges the request and assigns the mobile/web application a unique key which
will be used to receive the blockchain’s request later on.
12

2) The API uses a queue manager to store the request in a queue. The queue has a number
of transmission queues equivalent to the number of peers involved in the application. In
the example shown in the figure there are two companies participating in the network.
The two peers may be, for instance, one roaster and one producer/cooperative, hence the
queue manager in this case has two queues. In this case, examples of requests might be
as follows: enroll a farmer/roaster, query the history and details of a specific coffee batch
from the blockchain, or update the status and location of a shipment as it moves along the
supply chain.
3) Nodes receive and process the requests coming from the queue manager. In Fabric, a trans-
action proposal is first sent to multiple peers for endorsement. Endorsing peers simulate
transactions and sign those that are valid. Endorsers execute transactions to ensure their
correctness, while committers only verify endorsements and validate transactions prior to
committing valid blocks to the ledger. Hence, endorsers are a special kind of committing
peers as all peers can commit blocks to the distributed ledger.
4) After the proposer collects enough endorsements satisfying the endorsement policy, it
submits the transaction to be added to the Ordering Service. Note that, before any request
is processed, the Certificate Authority (CA) must authenticate a peer and approve the
request for that specific transaction.
5) The Ordering Service verifies the signatures and accepts endorsed transactions, arranges
transactions into a specific order, combine transactions into blocks and delivers the blocks
to the committing peers in the network, which in turn append the blocks to the ledger and
update it. It is the Orderer that decides the order of transactions and agrees it with the
orderers of other organizations.
6) The nodes then process the response to the original request and hold it in the distributed
memory-key database using the initial unique key.
7) In the end, the web/mobile app receives the response to its original request through the
REST API after querying the distributed ledger using the unique key.
Fantine.io is a Colombian startup that delivers unprecedented transparency to the coffee value
chain and allows farmers to transact directly with roasters. The need for this solution is critically
important for Colombia in particular, where coffee is often the only viable option for rural people
to escape the entrapments of illegal crop markets and guerilla groups.
13

V. CONCLUSION
This article starts with an introduction on the coffee value chain highlighting the challenges
the industry is facing, especially at its weakest ring, the producers. In the past decades, in fact,
coffee growers have seen their margins gradually squeeze and disappear. Borrowing concepts
from economics, the article introduces the concept of Commodity Fairness Index, which is novel
to the best of the authors’ knowledge, that can be used to measure quantitatively the imbalance
in any commodity value chain. Further, the paper calculates it in the specific case of Colombian
coffee, highlighting a very unbalanced industry and showing that there is room for improvement
in order to making it fairer and more balanced. The article also provides a number of ways that
such index could be improved; we leave these avenues to future work.
While this newly defined CFI index only helps in measuring quantitatively the inequality in a
commodity value chain, blockchain technology actually paves the way to improving it by creating
more equity and fairness for the growers in the value chain. This will in turn enable to realize
a more balanced and sustainable industry, allowing end consumers to feel better about guzzling
their caffeinated beverages, while simultaneously creating additional value in both producing
and consuming countries as well as at the consumer end.
REFERENCES
[1] L. F. Samper, D. Giovannucci and L. Marques Vieira, “The powerful role of intangibles in the coffee value chain,” WIPO
Economic Research Working Paper No. 39, Nov. 2017.
[2] L. F. Samper and X. F. Qui ˜
nones-Ruiz, “Towards a Balanced Sustainability Vision for the Coffee Industry,” MDPI Journal
of Resources, vol. 6, no. 17, Apr. 2017.
[3] Macrotrends, “Coffee Prices - 45 Year Historical Chart,” Available online at https://www.macrotrends.net/2535/
coffee-prices-historical-chart-data.
[4] L. Amado and F. Miatton, “Fueling the Revolution of the Coffee Value Chain: Fairness and Transparency in the Coffee
Industry through Blockchain and Machine Learning,” Fantine.io White Paper, Jun 2019.
[5] World Economic Forum, “Inclusive Deployment of Blockchain for Supply Chains: Part 1 Introduction,” White Paper, Apr.
2019.
[6] M. P. Caro, M. S. Ali, M. Vecchio and R. Giaffreda, “Blockchain-based traceability in agri-food supply chain management:
A practical implementation,” IoT Vertical and Topical Summit Agriculture, May 2018.
[7] K. Salah, N. Nizamuddin, R. Jayaraman and M. Omar, “Blockchain-based soybean traceability in agricultural supply
chain,” IEEE Access, vol. 7, pp. 73295-73305, May 2019.
[8] S. Wang, D. Li, Y. Zhang and J. Chen, “Smart Contract-Based Product Traceability System in the Supply Chain Scenario,”
IEEE Access, vol. 7, pp. 115122-115133, Aug. 2019.
[9] Provenance, “From shore to plate: Tracking tuna on the blockchain from catch to consumer,” Available at: https://www.
provenance.org/tracking-tuna-on- the-blockchain#pilot- phase-2.
14

[10] IBM Food Trust. “Digitizing the world’s food supply”. Available at: https://www.ibm.com/blockchain/solutions/food- trust.
[11] Hyperledger Fabric. https://www.hyperledger.org/projects/fabric.
[12] S. Ponte, “The Latte Revolution? Regulation, Markets and Consumption in the Global Coffee Chain,” World Development,
vol. 30 no. 7, Jan. 2002.
[13] L. G. Bell and P. Liberati, “Inequality Analysis: The Gini Index,” EasyPol vol.40, Food and Agriculture Organization
(FAO), Dec. 2006.
[14] D. J. Teece, “Business models, business strategy and innovation,” Elsevier Long Range Planning, vol. 43, no. 2-3, pp.
172-194, Apr. 2010.
[15] J. Garca, “Evoluci´
on de la distribuci´
on de las fincas cafeteras,” Federaci´
on Nacional de Cafeteros de Colombia, 2004,
Available: https://federaciondecafeteros.org/static/files/3.evolucionfincascafeteras.pdf.
[16] Enveritas and Carto, “Bringing Smallholder Coffee Farmers out of Poverty,” Dec. 2019, Available at: https://carto.com/
blog/enveritas-coffee-poverty-visualization.
[17] World Coffee Producers Forum, Final Declaration, Colombia, Jul. 2017. Available at: https://www.
worldcoffeeproducersforum.com/wp-content/uploads/2017/07/Declaracion-final- WCPF en-1.pdf.
[18] Hartman Group, Sustainability 2017 Report: https://store.hartman-group.com/content/Sustainability-2017-Overview.pdf.
15