=Paper=
{{Paper
|id=Vol-1481/paper8
|storemode=property
|title=Enabling Semantic Web for Precision Agriculture: a Showcase of the Project agriOpenLink
|pdfUrl=https://ceur-ws.org/Vol-1481/paper8.pdf
|volume=Vol-1481
|dblpUrl=https://dblp.org/rec/conf/i-semantics/TomicWHAD15
}}
==Enabling Semantic Web for Precision Agriculture: a Showcase of the Project agriOpenLink==
https://ceur-ws.org/Vol-1481/paper8.pdf
Enabling Semantic Web for Precision Agriculture:
a Showcase of the Project agriOpenLink
Dana Tomic Domagoj Drenjanac Wilifred Wöber
FTW FTW Universität für Bodenkultur
Donau-City Strasse 1 Donau-City Strasse 1 Gregor-Mendel-Strasse 33
Vienna, Austria Vienna, Austria Vienna, Austria
tomic@ieee.org drenjanac@ftw.at wilfried.woeber@boku.ac.at
Sandra Hörmann Wolfgang Auer
Josephinum Research MKW electronics GmbH
Rottenhauser Straße 1 Jutogasse 3
Wieselburg, Austria Weibern, Austria
sandra.hoermann@josephinum.at wolfgang.auer@mkwe.com
ABSTRACT
This paper describes the agriOpenLink approach towards
1. INTRODUCTION
triplification of the production data generated by heterogeneous 1.1 Precision Farming
agricultural equipment, for their easy integration and querying Precision agriculture stands for management practices and tools
within a common information context of various decision support that leverage advanced sensor, actuator and decision support
applications. The presented approach has been developed within technology to aid in optimization of the production processes
the running project agriOpenLink, which aims at improving [1][2].
agricultural production processes. Triplification is performed by
equipment-specific model adapters (plugins), which translate the The market for the precision agriculture equipment, both for the
equipment data into RDF triples. The plugins are realized by arable farming and the livestock farming is rapidly growing, and
means of semantic REST services, and triplification is a result of a the robotic and sensor systems are already indispensable at larger
workflow orchestrated "on-demand" by chaining services of farms. These are expected to be ubiquitously used in the years to
appropriate data sources. In the project, we have also come. Today, modern agricultural equipment such as
conceptualized, and are currently implementing a platform and autonomously driving tractors with their smart implements,
tools that facilitate easy plugin development and deployment. milking and feeding robots or animal monitoring systems,
Dealing with the problem of creating a common information aggregate many different types of sensors and collect many
context, the project designed a domain ontology taking into different sensor data. In many cases, these data are stored in local
account existing work, experimented with triplification of existing databases, e.g., imbedded within robots or accompanying
domain knowledge, and proposed an approach for the ontology controllers (PCs) that provide user applications. The raw sensor
maintenance. The agriOpenLink platform is being tested and data are often processed locally so as to provide users with a
demonstrated in the precision irrigation use case and the precision meaningful aggregated information for the decision support. Such
dairy farming use case. interpreted data is often presented to users in a tabular of
graphical form. Cloud-based solutions are being increasingly
offered, however the issues of data ownership or reuse by 3rd
Categories and Subject Descriptors parties are still subject of many controversies hampering the take-
H.4 [Information Systems Applications]: Miscellaneous off of these solutions.
Nevertheless, the equipment is just one source of the production
General Terms data. The farmers need to interact with a multitude of other
Design, Experimentation external information systems with the weather data, soil, weed,
crops, pest, animal breeds, feed, and other information. Although
Keywords the integration of data from different sources is a paramount
enabler for the decision support applications and for the
Triplification, ontology, semantic services production process optimization, it still remains a grand
challenge. Today it is often a farmer who has to collect data from
different sources e.g., in form of comma separated value (CSV)
files, and link them, and process them together. This is
particularly noticeable in the domain of precision livestock
farming where proprietary data models are often used, partially
because of a low acceptance of existing standards (e.g.,
ISOagriNET framework, www.isoagrinet.org) and their often
criticized inflexibility. As an answer to strong user needs,
26
emerging data integration solutions [3] typically also fall back to promises of the Semantic Web, the project agriOpenLink1
using proprietary data models, without attempting to solve the developed a semantic approach with a goal to establish a common
problem of data model interoperability. These approaches focus context for the integration of heterogeneous interfaces and data,
more on providing a single integrated user interface for the and to develop tools to facilitate in better decision making and
farmers, and less on establishing open interfaces for any new process optimization. This approach integrates three tasks: 1)
system that can be meaningfully connected. Therefore they often establishing a platform in which a common ontology-based model
lead to non-scalable solutions sensitive to changes in data formats can be created and maintained to reflect the common data
that are used on the interfaces of heterogeneous integrated integration possibilities and needs, 2) designing tools to ease the
sources. development of the model adapters – the equipment-specific or
data-source-specific triplifiers – that translates process data into
1.2 Semantic Web for Precision Farming RDF triples based on the concepts of the common ontology, and
The goal of Semantic Web is to use Web as an information 3) designing a run time environment in which data translation
management platform [4]. The standards of the Semantic Web (tripligication) is performed on demand as a result of execution of
enable unified modelling of data and metadata, so that they could queries. As these RDF data relate to the common ontology (the
be incrementally integrated within transparent knowledge knowledge context) they can be easily integrated and queried
structures within which concepts can be disambiguated and together.
interrelated, and data easily linked to their defining schemas. The
While agreeing on a common model (a defacto standard) is a
standards, tools and practices of Semantic Web have reached
prominent requirements in the agricultural sector, it is not
considerable maturity, and are used for the management of both
possible to aspire model completeness. Therefore, the model
open data (freely accessible) and enterprise data (accessible to
extension or change has to be technologically easy to implement
registered users) in many different sectors and applications [5][6].
and propagate. This requirement makes semantic technology the
These tools support the whole lifecycle of creating and
technology of choice for creating and maintaining a standard
maintaining linked data, which includes 1) extraction of
model. Ontology-based domain knowledge is easy to extend and
structured data from other sources, 2) data authoring or creation
interlink with already established concepts. That is why
via “triplification”, enrichment, interlinking and fusing and 3)
agriOpenLink focus on tools for the ontology-based lifecycle
data maintenance in a repository. As a result of the Linked Data
management of the schema and data. Accordingly adding a new
community effort the Open Linked Data Cloud today include
data source or changing formats does not result in a detrimental
huge number of data sets (ontologies, vocabularies, … and real
complexity or a high implementation cost.
data) with geographic information, publications, user generated
content, government, life sciences, and cross-domain. The The agriOpenLink approach also aims to account for specificities
enterprise systems can re-use established semantic descriptions of data sources in agricultural production environment and special
and uniquely defined resources and benefit from the linked data requirements on structuring them. While the applicability of the
concepts and the shared semantics. relational database schema translation approaches is high for the
owners of large databases, this does not apply for the equipment
The languages and tools for translation of data from different
as a source of production data, because data stored within the
formats into RDF are important components of existing Semantic
equipment database first needs to be processed (e.g., averaged
Web tool chains. Several approaches have already been proposed,
within a specific interval, etc.) to be useful for sharing. Therefore,
demonstrated and offered as standards. For the translation of
agriOpenLink uses plugin approach and application specific
relational databases the W3C defined R2RML [7]. The RML
translation of data into RDF by means of semantic REST services.
approach [8] defines a language that extends the R2RML standard
particularly addressing concurrent translation of many related
sources. These approaches are very well suited to be used by 2.2 The Realization
owners of large databases that have already accumulated huge 2.2.1 Plugins as Triplifiers
amounts of data. The agriOpenLink platform offers tools and procedures to
Within the agricultural domain the Semantic Web approach translate the agricultural data from the precision agriculture
already inspired a number of solutions in the research spectrum, equipment or other agricultural information sources into their
e.g., [9-13] and is also embraced by the Food and Agriculture semantic RDF-based form, in which they can be interlinked and
Organization of the United Nations (FAO; http://aims.fao.org) in jointly queried. We adopted a plugin-based approach where each
their global initiatives for agricultural information management source of a relevant information is wrapped within a so called
systems (AIMS), AGROVOC vocabulary, and agricultural plugin acting as a model adapter - a triplefier. The triplification is
ontology service [14]. Also in the agricultural domain there are realized as service-based business logic. More specifically a
many different sources of information, and actors who currently plugin is a software component which publishes semantic REST
maintain these data for the farmers and are opening them via services, similar to the SADI Web Services [17] that consume and
APIs, e.g. as described in [15], or can be found on [16]. produce RDF triples. A plugin acts as a model mediator – it
translates data from the internal model of the device into instances
of particular class of the ontology decorated with specific
2. THE AGRIOPENLINK APPROACH properties. Accordingly its function is to open and provide for
2.1 The Motivation further linking RDF data from the devices. An important
Motivated with the shortcomings of the existing data integration component of our plugin approach is the development
approaches in commercial precision farming solutions and the
1
www.agriopenlink.com
27
environment which offers plugin skeleton code and basic We experimented with the ways of encoding the expert
functionality to aid in the plugin implementation. knowledge in form of defined classes, SPARQL queries and
semantic services. To this aim the Ontology Editor offers a user
The decision to create a specific plugin platform and not use
interface to create defined classes with restrictions on properties
existing Internet of things (IOT) frameworks such as [18], and to
as a basis for classification. An example is a class that restricts the
encapsulate translation of the data into RDF within semantic
activity property to define an animal which may be potentially
services is based on specific requirements and current constraints
lame. The Query Editor offers a user interface for the creation of
existing within the agriculture sector. First of all, we aimed at a
SPARQL queries. We are currently implementing a solution with
compact solution with a minimum set of functionality. The goal is
which any query can be also translated into 1) a dynamically
to bring the benefits of the semantics and ontology to the involved
deployed service (which controls invocation of other services) and
actors (equipment vendors/ 3rd party application providers), but
2) an API which can be programmatically included in any 3 rd
hide as much as possible the complexity of the semantic
party application that wants to start such query.
representation. To this aim the ontology is maintained via the
platform, and the plugin development environment automates 2.2.2 The Domain Ontology
some of the complexity of semantic data processing and semantic Particularly the creation of the domain ontology for the dairy
REST service creation, for the plugin developers. Accordingly, farming scenario was challenging, due to not much of existing
agriOpenLink aims at scalable integration of any data interface work in this domain. In the process of DFO engineering for the
and data format within a common information and knowledge domains of milk quality, feeding, breeding, fertility, we analyzed
context in which data can easily be related to and combined with both the ISOagriNET standard dictionaries [16] and the schemas
other data. of the milking robots, concentrate feeder, and heat and activity
While many agricultural equipment use relational databases to monitoring equipment. We created the first version of the
organize its internal data, our triplification approach is not based ontology in the OWL ontology editor Protégé. It addresses the
on the approaches and tools for data model translation such as scenario in which a linked data set is created for a herd at the farm
[7][8]. The reasons for this are twofold. Firstly, the raw data as a basis for herd management applications. The herd linked data
generated by the equipment is very often quite sensitive and must set includes the farm core data, i.e., the animal registration
first be processed into interpreted data that can be offered to information, and is enriched with the data continuously coming
farmers. The raw data is often considered as being owned by the from different farm equipment and external sources, including the
equipment vendors. Accordingly, the amount of interpreted milk quality, feed, activity, fertility, and health information. We
information is lower as compared to the raw data. The information started with a small set of OWL classes including: Animal, Farm,
offered to the farmer is sometimes not in the database, but is Farmer, Equipment, Organization and the object properties to
calculated and only presented on the user interface. Secondly, the reflect the relationships between instances of these classes such as
engineers implementing the internal logic of this equipment, and parent-child relationship between Animals, similar to the
accordingly the plugins, do not have to learn semantic-heavy tools ontologies [11][21]. While these existing ontologies provided an
but can program the business logic of translation completing the important input, only the latter one has been recently published.
plugin skeleton. During the ontology creation it became clear that standard
The project agriOpenLink is demonstrating the use of the platform ISOagriNET dictionaries need to be available in the linked data
in two scenarios: (1) in the precision irrigation scenario we focus format in order to reference them in a formal way. Consequently,
on triplifying weather, soil and sensor data and expressing the we defined their “triplification” as a goal for the 2nd phase of our
transpiration model knowledge in form of semantic services [19]. DFO engineering task. In finalizing the 1st phase we revisited the
In the precision dairy farming scenario we triplify data from the requirements of our initial herd linked data scenario and re-
farm robots and farm systems and experiment with expert focused the DFO engineering on the animal state diagnoses
knowledge for animal state diagnosis expressed as SPARQL scenario, in which WC3 SPARQL queries coming from Decision
queries [20]. Figure 1 illustrates the platform architecture in a Support Systems trigger search, access, interlinking and filtering
precision dairy farming. A plugin server that implements HTTP of semantic data available via Plugin Services. We reduced our
REST interfaces is a central component of the local system that DFO to the properties that belong to a core data set, and that can
initiates and run different plugins on the farm. The agriOpenLink be extracted from equipment and external sources. DFO further
platform also includes a plugin development environment in includes defined classes that are used by Plugin Services. They
which plugins can be designed and built into dynamically specify what properties can be extracted from each specific
loadable components. Plugin, i.e., specific type of the equipment.
2.2.3 Translating ISOagriNET Data Dictionaries
To obtain an RDF/OWL model of the standard dictionaries for
livestock farming we developed a triplifyer tool that follows the
best practices for creating the linked data [6]. Our triplifier tool
can also transform standard exchange data files into linked data.
The triplification of the standard schema was described in [22].
To test the procedure we triplified all dictionaries available via
http://ian.lkv-nrw.de as well as exemplary data files. The
translated dictionary for the year 2003 is available at
www.agriopenlink.com/ADR2003. We are currently working
towards completing and opening the ontology platform with
ontology browsing and editing GUI and a SPARQL interface.
Figure 1. The agriOpenLink Platform (dairy farming 28
scenario)
.
3. CONCLUSIONS [8] Dimou, A., Vander Sande, M, Colpaert, P, Verborgh, R,
The presented agriOpenLink platform is a work-in-progress Mannens, E. and R. Van de Walle. RML: A Generic
solution that was designed by strongly focusing on data created Language for Integrated RDF Mappings of Heterogeneous
within the agricultural production environment, on the data Data. Proceedings of the 7th Workshop on Linked Data on
formats that are currently used in these environments, and on the the Web, 2014.
benefits of translating production data into RDF for their easy [9] Eckartz, S, Verhoosel, J, Folmer E and E Somers 2013.
integration and integrated querying. Semantics for Smart Dairy Farming: a milk production
registration standard v1.0, Report, available at
Today farmers need to use data from many different data sources
http://www.smartdairyfarming.nl, last accessed Aug 2015.
and to integrate them in a meaningful way. Very often relevant
data sources, such as agricultural equipment, do not offer open [10] Chaplinskyy, Y. and O. Subbotina, 2013. Ontology-driven
data interfaces, so farmers need to either manually verify data or advice and decision-making within knowledge,
dump data into csv files and process them in some common dissemination for extension, HAICTA 2013.
purpose tools. Also, some data interfaces do not comply with the [11] Gao Yongchun, 2005. The application of Web Ontology
existing ISOagriNET standard, and often it takes a long time to Language for information sharing in the dairy industry, PhD
introduce new data properties pertaining to innovative new Thesis, McGill Univ. Montreal. 2005.
sensors and systems into this standard.
[12] Athanasiadis IN, Rizzoli AE, Janssen S, Andersen E, Villa F.
The contribution of the agriOpenLink solution is in offering an 2009. Ontology for Seamless Integration of Agricultural Data
integrated approach and a platform which can support 1) and Models. MTSR 2009; Milan, Italy, October 1-2, 2009,
maintenance of the domain knowledge (ontology) in the 282-293.
repository with the SPARQL and user interfaces for ontology [13] Grimnes MK, Abufouda, M and A Schröder. 2012, Semantic
community-based editing, 2) adding of new devices by means of Integration through Linked Data in the iGreen project. GIL
plugins that publish RDF data complying to the ontology, and Jahrestagung, volume 194 of LNI, page 107-110.GI,(2012)
which can be easily created in a plugin development environment,
and, 3) querying of RDF data on-demand, where the resulting data [14] Lauser, B., Sini, M., Liang, A., Keizer, J. and S. Katz,
can be stored for further publishing and querying. We have Stephen, 2006. From AGROVOC to the Agricultural
implemented and demonstrated core functionalities of the Ontology Service / Concept Server, available at
platform in a demonstrator and are currently focusing on http://aims.fao.org
deployment on farms, as well as on completing and opening the [15] Haezebrouck, T-P, Emeric, E, Sine M, API-AGRO: An Open
agriOpenLink ontology maintaining platform for experimentation. Data and Open API platform to promote interoperability
standards for Farm Services and Ag Web Applications,
4. ACKNOWLEDGMENTS EFITA 2015, June 29-July 2, Poznan, 2015.
The work presented here is partially funded by the Austrian [16] LKV, 2015 ‘ISOagriNet, http://ian.lkv-nrw.de/index.php,
Research Agency (FFG) within the project agriOpenLink http://ian.lkv-nrw.de, http://webapp.lkv-
(research fund 898398). nrw.de/AdedDataDictionary, last accessed Aug 2015
[17] Wilkinson, M.D., McCarthy, E.L., Vandervalk, B.P.,
5. REFERENCES Withers, D., Kawas, E.A., Samadian, 2010. S.: SADI,
[1] McBratney, A., Whelan, B., Ancev, T., 2005. Future SHARE, and the in silico scientific method. ;BMC
Directions of Precision Agriculture. Precision Agriculture, 6, Bioinformatics(2010)7-7
7-23. [18] FIWARE Backend Device Management – IDAS,
[2] Bewley, J.M. 2013. Exciting Dairy Breakthroughs: Science http://catalogue.fiware.org/enablers/backend-device-
Fiction or Precision Dairy Farming? Proceedings of the management-idas, last accessed Aug 2015
Precision Dairy Conf., Rochester, MI, June 26-27, 2013. [19] Wöber, W.; Supper, G.; Aschauer, C.; Gronauer, A.; Tomic,
[3] Elite, 2014 ‘Drei Programme für die Organisation Ihrer D.; Hörmann, S., Entwicklung eines auf semantischer
Herde‘, in German, Elite Magazin 6/14, 2014. Technologie basierenden Analysesystems zur Überwachung
der Wasserversorgung von landwirtschaftlichen Nutzflächen,
[4] W3C Semantic Web,
GIL Tagung 2015, Available at http://www.gil-
http://www.w3.org/standards/semanticweb/, last accessed:
net.de/Publikationen/27_201.pdf, Last accessed June 2015
Aug 2015.
[20] Tomic SDK, Hoermann S., Handler F., Wöber W., Otte, M.,
[5] Auer S., 2014. Introduction to LOD2, in Linked Open Data
and W. Auer, 2014. agriOpenLink: Semantic Services for
– Creating Knowledge Out of Interlinked Data, 1-17,
Adaptive Processes in Livestock Farming, AgEng 2014.
Springer, 2014.
[21] Seamless , 2015. http://ontologies.seamless-
[6] Mezaour, A-D., Van Nuffelen, B., and C. Blaschke. 2014.
ip.org/livestock.owl, last accessed Aug 2015
Building Enterprise Ready Applications Using Linked Open
Data, in Linked Open Data – Creating Knowledge Out of [22] Tomic, D., Drenjanac, D., Hoermann S., Auer, W,
Interlinked Data, Springer, 2014. Experience with Creating and Maintaining Ontology for
Precision Dairy Farming: A Showcase of agriOpenLink,
[7] W3C Semantic Web, R2RML: RDB to RDF Mapping
EFITA 2015, June 29-July 2, Poznan, 2015.
Language, (2012) http://www.w3.org/TR/r2rml/, last
accessed Aug 2015.
29