ICT-AGRI Call 2 Mid-term Project Report: SILF: Smart Integrated Livestock Farming

Frank Oudshoorn, Thomas Bartzanas, Nicholas Holden, Mikko Järvinen, Allan Jensen, Ludwig Lauwers, Leen Vandaele, Tim Van De Gucht, Annelies Van Nuffel, Wouter Saeys, Claus Sorensen

Onderzoeksoutput: Boek/rapportRapport

3 Downloads (Pure)


Available databases of relevance for development of an IoT data management platform for livestock farming
were identified through a survey in the five partner countries Belgium, Denmark, Finland, Greece and Ireland.
The results indicate multiple available databases within each country, with data of relevance for livestock health
and welfare monitoring, both in the type of on‐line as well as non on‐line configurations. The results will form
the basis for combining data into a dedicated farm journal. As part of such a farm journal input have been
provided in the form of an energy audit protocol.
In terms of analyzing stakeholder involvement, identification of stakeholders was initiated and focusing on
two technologies, lameness and energy consumption and for two processes: 1) the design of the technology, 2)
the assessment of sustainability of the technology, focusing on different phases of the life‐cycle: Usage,
Maintenance, Implementation, and Production (construction). Important stakeholder for the design process
includes the farmers and the advisors. For, implementation and production phase within design and
assessment, the farm owner and the manufacturing industry were rated high. In continuation of the
stakeholder analysis, survey consisting of questions to farmers about general farm issues and specifically about
lameness was initiated. The choice experiment consists of some questions (referred to as choice sets) where
the farmer needs to choose between three alternative systems and pick the one they prefer most. Responses
are currently being collected.
Preliminary experiments with accelerometers have been carried out, to try to identify parameters and
classifiers of lameness. 5 collars were fitted with 2 accelerometers each, the accelerometers were configured
to sample at the maximum frequency (200Hz) rate to minimize the risk of missing cues in the data. This setup
had the purpose of assessing the sensor positions influence on the ability to detect lameness. In the first
experiment, there seem to be some difference in the movement/activity level of the lame cow during resting
periods. However there is no obvious detection criterion to extract from the data, that wouldn’t give a high
false positive rate. To make any definite conclusion more data is needed, along with validation data. Other
developments within the area of testing lameness detection systems include that the the Stepmetrix system is
being discarded, the Gaitwise system experienced malfunctions but is in repair. Demonstration farms in Finland,
Belgium and Ireland search for alternative accelerometers or accelerometer based activity sensors. However,
algorithms for accelerometers are being further developed on experimental basis.
A list of key environmental indicators was identified. These are currently being reviewed to propose a
shortlist for wider discussion. The indicators include categories within energy, nutrient use, soil/land issues,
biodiversity, water, carbon footprint and economy. The indicators form the basis for a farm based LCA where
economic drivers can be connected. The next step will include identification of a number of farms in DK and IRE
to test the LCA and relate the scoring to detection system data.
System analysis has been performed by indicating the identified stakeholders. A web has been made that
represents the relations between different actors. This web will be updated continually with information about
these relationships, as well as costs and benefits for different actors. A similar approach to the SILF web was
used to make a farmer web that summarizes the different effects and economic consequences of lameness and
their mutual relations. This web will be further updated to clarify the known and unknown costs in literature.
Deviations from the original project plan involve that the partner WebsTech went into bankruptcy
preventing the provision of the test sites with accelerometers. The Stepmetrix system is not working, and the
Belgian partner doesn’t want to go on with this. Furthermore, the Gaitwise system has been repaired after the
cold experience in Finland and is now in a new configuration to be put up in Belgium and Finland but it is not
viewed as a usable reference system. Suggestions for alternations to the project are drafted.
Oorspronkelijke taalNederlands
Aantal pagina’s16
PublicatiestatusGepubliceerd - 1-sep-2014


  • T420-landbouwtechnologie-landbouwmachines-boerderijbouw

Dit citeren