We build on past work proposing the use of Linked Pedigrees in a semantic technology framework to propose the use of blockchain technology to solve part of the trust issues in the agri-food supply chain.
2. Outline
• Information integration in the agrifood supply chain
• Semantic web technologies
• Semantic Web technologies in the Food System: Linked
Pedigrees
• Some limitations
• Blockchain technologies
• Integrating Semantics and the Blockchain - some initial
thoughts
4. The Food Supply Chain
• From Farm to Fork
• The agri-food system
includes much more
• More and more parts of
this supply chain and
agri-food system are
leaving digital trace … in
James Scott’s terms
becoming more “legible”.
5. Data - Information -
Knowledge
• The food supply chain involves hundred of
actors, thousands of processes, millions of
products and (potentially) billions of data points!
• Children believe milk comes from supermarkets!
• Too much or too little data?
• Why do we need it?
6. Characteristics of Supply Chain
• Large numbers of participants
• Heterogeneity of participants
• Huge variety in ICT uptake
• Poor information flow (need to know attitude)
• Only one up, one down data flow
• Solved by regulation and certification
7. Food supply chain is …
• A highly heterogenous loosely coupled large-scale
network of entities with variable but largely minimal
degrees of communication and trust between the actors
8. Drivers for Data Integration
• Need for transparency - tracking and tracing
• Desire for food awareness - on the part of
consumers, but not only
• Regulatory pressure - e.g.EU Regulation
1169/2011
• New business opportunities ….
9. Food Crises and Scandals
• Major driver for greater data integration (whether
open or closed).
• E. Coli in Germany in 2011 - spanish growers
lost over €200M
• Horsemeat scandal across Europe in 2013 -
impact very great on some supermarkets
10. Lack of Data Integration
• Both scandals suffered from lack of data and data
integration
• E. Coli - who affected? what purchased? where?
when? and who participated in the supply chain
• Horsemeat - six months for Irish FSA to map the
supply chain network
• Need for greater supply chain transparency =
need for data integration
11. Tracking, tracing and
Visibility
• Core demand is to make tracking and tracing
easy, AKA supply chain “visibility”
• “Visibility is the ability to know exactly where
things are at any point in time, or where they
have been, and why” — GS1
• Major challenge
http://www.gs1.org/docs/GS1_SupplyChainVisibility_WhitePaper.pdf
15. Key Features of Semantic
Technologies
• Unique identifiers (URIs) — enables consistency and data
accretion
• Common vocabularies/ontologies/data schemata — creates
a tendency towards standardisation WITHOUT losing
flexibility
• Linking and mappings - create a natural space for new
knowledge and data integration
• Logical rigour — rules for validation and quality control can
be written
16. AKTive Food (2005)
• A bit of history
• Paper on “Semantic Web based knowledge
conduits for the Organic Food Industry”
• Centred on decision support for a restaurant
based on data crawled from semantic web
marked up websites of food producers
• Nice vision …. but no implementation
17. Linked Pedigrees
• Based on “pedigree” concept common in
pharmaceutical industry - an audit trail which record
path of ownership
• Based on GS1 standards (pedigree standard +
EPCIS)
• “Linked pedigrees” use semantic web/linked data
principles
• Involves formalisation of EPCIS standard in two
ontologies
18. Linked Pedigrees
• Datasets described and accessed using linked data principles.
• Encapsulate the knowledge required to trace and track products in
supply chains on a Web scale.
• Facilitate the interlinking of a variety of related and relevant data,
i.e., GS1 product master data with event data PLUS other data
outside the GS1 system.
• Based on a domain independent data model for the sharing of
knowledge among Semantic Web/Linked data aware systems
deployed for the tracking, tracing and data capture.
• Product knowledge shared among partners as products physically
flow downstream or upstream in the supply chain.
20. Linked Pedigrees and EPCIS
• Formalisations of:
• EEM - The EPCIS Event Model
• CBV - Core Business Vocabulary
• This allows the representation of
EPCIS events on the Web of Data
• This enables sharing and
traceability of information
• Tracking of inconsistencies
21. Socio-technical Limitations
• Heterogeneity of the food system - so many
different actors, in size and shape
• Continuous changes - actors entering and
leaving the market
• Lack of trust - actors (farmers, food producers)
do not trust overarching systems
• Cost - margins in the food system are very tight
22. Key Problems
• Any form of centralised
data runs into data control
issues. Who know what?
(cf. Uber as an example)
• If each actor must keep
their triple store up and
running - data access
issues (important in food
crises)
24. What is the blockchain?
• A file called The Blockchain is spread over
millions of machines
• Which use proof of work and byzantine
consensus
• To provide a set of chained hashes and digital
signatures
• To create an unforgeable record of …. (e.g.) who
owns how much bitcoin
25. Blockchain - another
definition
• “A blockchain is a magic computer that anyone
can upload programs to and leave the programs
to self-execute, where the current and all
previous states of every program are always
publicly visible, and which carries a very strong
cryptoeconomically secured guarantee that
programs running on the chain will continue to
execute in exactly the way that the blockchain
protocol specifies.” — Vitalik Buterin (founder of
Ethereum)
26. Blockchain
• Developed originally as part of Bitcoin
• Provides underlying distributed ledger for Bitcoin
• HOWEVER, quite separate from Bitcoin and has
potentially many other uses
• Lots of eager uptake with many startups being
founded around this technology (Ethereum,
Bitshares, Helloblock, Ripple Labs etc.)
27. Blockchain in the supply
chain
• Not my idea! Other people have thought of this!
• Startup provenance.org wants to use the blockchain to
“tell a story” about a product from producer to end
consumer. Currently focussing on certification data!
• Still working on on what data to represent ….
28. Semantic Blockchains
• Concept: Construct an architecture where some or all of
the data involved in Linked Pedigree is held on the block
chain
• Result:
• Distributed database would resolve some trust issues
• Guarantee of continuous uptime (so if an actor
disappears, their data is still accessible)
• Rules can be written as to who has access to data
using specific governance algorithms
29. Step 1: Basic Usage
Eliminate the
problems of data
centralisation
30. Step 2: More advanced Usage
Guarantee
accessibility of data
now and in future
31. Other potential
Consequences
• Disintermediation of GS1 for product data
• Product data, tracking and tracing and supply
chain visibility at very low cost. This could be
very important for small scale producers/
developing country producers
• Standardising supply chain data schemata/
ontologies by the back door
32. Conclusions
• We have argued for the importance of data
integration in the agrifood supply chain
• We showed the applicability of semantic
technologies in the supply chain and introduced
the concept of “linked pedigrees”.
• We then suggest that blockchain technologies
could further improve this technology stack and
solve some problems e.g.lack of trust in
centralised data control
33. Thanks
• Monika Solanki for all the technical work on
Linked Pedigrees, EEM, CBV and much else
• Vinay Gupta for conversations leading to the
Semantic Blockchain ideas
• Jessie Baker for explaining provenance.org
34. References
• Christopher Brewster, Hugh Glaser, and Barny Haughton. “AKTive Food: Semantic Web based knowledge conduits for the
Organic Food Industry.” In Proceedings of the ISWC Worskshop Semantic Web Case Studies and Best Practice for
eBusiness (SWCASE 05) , 4th International Semantic Web Conference, 7 November (Galway, Ireland, 2005). URL http://
www.cbrewster.com/papers/Brewster_SWCASE.pdf
• Monika Solanki and Christopher Brewster. “OntoPedigree: A content ontology design pattern for traceability knowledge
representation in supply chains.” Semantic Web – Interoperability, Usability, Applicability (2015). URL http://goo.gl/OdUPg0
• Monika Solanki and Christopher Brewster. “Enhancing visibility in EPCIS governing Agri-food Supply Chains via Linked
Pedigrees.” International Journal on Se- mantic Web and Information Systems 10 (2014). (Impact Factor 0.393), URL http://
www.ijswis.org/?q=node/52
• Monika Solanki and Christopher Brewster. “Consuming Linked data in Supply Chains: Enabling data visibility via Linked
Pedigrees.” In Proceedings of the Fourth Inter- national Workshop on Consuming Linked Data (COLD2013), held at the
Interna- tional Semantic Web Conference (ISWC 2013), 21-25 October 2013 (2013). URL http://www.cbrewster.com/papers/
Solanki_COLD13.pdf
• Monika Solanki and Christopher Brewster. “Representing Supply Chain Events on the Web of Data.” In Proceedings of the 3rd
International Workshop on Detection, Represen- tation, and Exploitation of Events in the Semantic Web (DeRiVE 2013), , held
at the International Semantic Web Conference (ISWC 2013), 21-25 October 2013 (2013). URL http://www.cbrewster.com/
papers/Solanki_DeRiVE13.pdf
• Monika Solanki and Christopher Brewster. “EPCIS event based traceability in pharmaceu- tical supply chains via automated
generation of linked pedigrees.” In International Semantic Web Conference 2014 (ISWC 2014) (Rivo di Garda, 2014). (ISWC
accep- tance rate: 21.1% 180 full submissions, 29 accepted, 9 conditionally accepted), URL http://www.cbrewster.com/
papers/Solanki_ISWC14.pdf