Abstract
Poultry and pig meat belongs to the most produced and consumed meat in the EU, being 13 million and 22,1 million of tons respectively in 2022. The production often occurs in some of the most intensive farming systems characterized by high stocking densities and indoor rearing conditions. Despite a decrease in the number of farms over the last years, there has also been a tendency of an increase in the average livestock farm size and a need for more external employees. Simultaneously topics such as antibiotic consumption reduction, animal welfare and environmental impact are expected to be addressed. Automated monitoring systems of the environmental conditions and animal behavior linked with productivity data will be a key element to identify and correct suboptimal housing conditions. A lack of synergy between technology developers and producers/animal health experts has led to the availability of many sensors but a lack of predetermined target values and decision support systems allowing data-driven decision-making. Moreover, both internal (technical specificities of the installations, occupancy and age, etc.) and external factors (wind, rain, etc.) determine the final stable climate and therefore make it crucial to combine all of these data for correlation and causation analysis so actionable insights are generated.
Therefore, the SENSORS PP project aims to:
Test and implement real-time negative pressure sensors applicable under stable conditions with a cloud-gateway connectivity
Apply natural language processing to digitize stable cards (mortality, medication treatments, water and feed-uptake)
Develop image analysis algorithms for pigs and poultry to allow automated analyses of animal activity, distribution and social behavior. These are early indicators of health or welfare issues.
Grow the data-base of environmental parameters and behavioral indicators in pig and poultry houses under different ventilation conditions from farms with and without climate related health or productivity problems so the environmental target values and thresholds can be refined in relation with the varying conditions.
Aggregate and process heterogeneous sensor data, external weather conditions and production parameters ingested into an IoT platform, enabling automated data analytics.
Built an interactive visualization information platform as a decision support system and early warnings method for the users.
Our experience has shown that in many farms the environmental conditions are not optimal. Monitoring and correcting environmental conditions of animals based on real time large data sets can contribute to a great extent to animal health, limiting treatment incidences, increasing productivity and improving welfare. By creating an entire monitoring toolbox composed of the hardware and software, applicable in poultry and pig houses, installed on a permanent or nomad basis and with an interactive user interface we aim to increase the implementation of such practices. The automated interpretations of the sensor data in correlation with animal behavior, productivity data and external factors should allow us to screen more farms in a less time-consuming matter and in a way that the installation of the devices can be done in cooperation with technicians. This will also mean that the implementation becomes cheaper and that reevaluation of corrective actions less restricted by price.
Keywords
Housing environment, AI, sensor technology, welfare, antibiotic reduction
Project coordinator
Dr Sjouke Van Poucke, SYN+BV, Belgium
Partners
- HUNGARY: Dr Julia Pinter, En-Co software
- GERMANY: Prof Bodo Kraft, FH Aachen
Funding institutions
- VLAIO, Belgium
- National Research, Development and Innovation Office, Hungary
- Federal Ministry of Food and Agriculture (BMEL)
- represented by the Federal Office for Agriculture and Food (BLE)
Project duration
36 Months
Abstract
Poultry and pig meat belongs to the most produced and consumed meat in the EU, being 13 million and 22,1 million of tons respectively in 2022. The production often occurs in some of the most intensive farming systems characterized by high stocking densities and indoor rearing conditions. Despite a decrease in the number of farms over the last years, there has also been a tendency of an increase in the average livestock farm size and a need for more external employees. Simultaneously topics such as antibiotic consumption reduction, animal welfare and environmental impact are expected to be addressed. Automated monitoring systems of the environmental conditions and animal behavior linked with productivity data will be a key element to identify and correct suboptimal housing conditions. A lack of synergy between technology developers and producers/animal health experts has led to the availability of many sensors but a lack of predetermined target values and decision support systems allowing data-driven decision-making. Moreover, both internal (technical specificities of the installations, occupancy and age, etc.) and external factors (wind, rain, etc.) determine the final stable climate and therefore make it crucial to combine all of these data for correlation and causation analysis so actionable insights are generated.
Therefore, the SENSORS PP project aims to:
Test and implement real-time negative pressure sensors applicable under stable conditions with a cloud-gateway connectivity
Apply natural language processing to digitize stable cards (mortality, medication treatments, water and feed-uptake)
Develop image analysis algorithms for pigs and poultry to allow automated analyses of animal activity, distribution and social behavior. These are early indicators of health or welfare issues.
Grow the data-base of environmental parameters and behavioral indicators in pig and poultry houses under different ventilation conditions from farms with and without climate related health or productivity problems so the environmental target values and thresholds can be refined in relation with the varying conditions.
Aggregate and process heterogeneous sensor data, external weather conditions and production parameters ingested into an IoT platform, enabling automated data analytics.
Built an interactive visualization information platform as a decision support system and early warnings method for the users.
Our experience has shown that in many farms the environmental conditions are not optimal. Monitoring and correcting environmental conditions of animals based on real time large data sets can contribute to a great extent to animal health, limiting treatment incidences, increasing productivity and improving welfare. By creating an entire monitoring toolbox composed of the hardware and software, applicable in poultry and pig houses, installed on a permanent or nomad basis and with an interactive user interface we aim to increase the implementation of such practices. The automated interpretations of the sensor data in correlation with animal behavior, productivity data and external factors should allow us to screen more farms in a less time-consuming matter and in a way that the installation of the devices can be done in cooperation with technicians. This will also mean that the implementation becomes cheaper and that reevaluation of corrective actions less restricted by price.
Keywords
Housing environment, AI, sensor technology, welfare, antibiotic reduction
Project coordinator
Dr Sjouke Van Poucke, SYN+BV, Belgium
Partners
Funding institutions
Project duration
36 Months