Achieving the UN sustainable development goal of a “world with zero hunger” by 2030 will require being more productive, efficient, sustainable, inclusive, transparent and resilient. This objective requires an urgent transformation of the current system of agriculture, livestock and food in general.
Industry 4.0 is transforming many industries with disruptive technologies like #Blockchain, Internet of Things (#IoT), Artificial Intelligence (#AI). In the agricultural and food sector, the spread of mobile technologies, IoT and Edge computing are already improving the access of small producers to innovative developments that improve their operations.
At #54cuatro we are convinced that the great challenge of companies like ours is to democratize access to these technologies that until recently were exclusive to large corporations.
In the world there are different solutions that allow monitoring all kinds of elements and, of course, animals. Animals of all types can be monitored, from cows, bulls, sheep, horses, including wild or water animals.
Just as there is communication between industrial equipment of the #M2M (machine to machine) type, we take on this challenge of putting together a communications system that we call #A2M (Animal to Machine).
Our methodology, unlike standard products, implies a custom development taking into account specific problems.
In this note we will develop the approach used for the Buffalo Monitoring project in the province of Chaco, Argentina.
Problem to solve
The situation that we found has to do with economic losses of about USD 300,000 due to the failure to locate the animals and the failure to detect heat, which meant that the Bufalas were not pregnant. The additional complexity came from the fact that the field is 6,000 hectares.
Monitoring on small, bounded farms is simple, but given so much terrain we had to change the focus of the project. The first thing we did was investigate the behavior of the bubalino. Thanks to different entities such as the International Buffalo Federation, we detected the following patterns:
• The bubalino has 150 sweat glands per cm2, unlike the cow that has 1,500. This means that it needs water to cool itself. This information helps us detect frequently located areas based on temperature.
• It lives on average 25 years unlike the cow that lives 10, and can give 16 calves against 6 that the cow gives. This marks the importance of locating the females to avoid losing heat cycles.
• When the Buffalo is in heat, it allows itself to be chased by the male or allows it to rest on the rump. We can detect symptoms of heat taking into account the behavior of their movements.
• The sick or life-threatening animal moves away from the group. This is important to control the cause of death and recover the sensors.
With these patterns we begin to design 3 things. On the one hand, the network coverage to detect the position of the animal. Secondly, the type of sensor, given that because of how the animal behaves, we could not use a common sensor because it would not last so long submerged, because the animal would rub it against trees to remove it, etc. And on the other hand, reporting patterns that allow us to detect location, possible heat conditions, disease, etc.
We design the sensor based on behavior. What we did was test designs on 3D models.
For connectivity we install:
• 3 communication masts at 3 full winds, 36 meters high, anti-rotor star, beacon. Civil work: high anchorages, field protection fence with doors, and there we set up 3 LoRa Gateways.
• Ubiquiti IP transport radio links and Mikrotik PoE routers in 10U outdoor cabinets, autonomous using solar energy (100ah panels and batteries).
Each animal was transformed into a transmission node. We use the geolocation platform developed by Odea to determine the positioning and cross the GPS data with the Ear Tag data that contains:
- UID Stick
- ID Posicionamiento
- Birth date
Additionally, we incorporated other types of Datapoints that were of interest to us, such as climatic factors and health schedules.
Finally, to reduce the detection times of the animal’s state, we adopted a drone equipped with a flight plan, a multispectral flight plan, and a wide visual range, thanks to the people of Runco who helped us find the best equipment for what we needed.
With the implementation of our solution, the platform will be providing insights from each node, which will nourish our data catalogs and allow us to adjust pattern detection algorithms. Those detected patterns should enable some key results:
- Find each animal
- Detect heat signals
- Understand the behavior according to temperature, humidity, rainfall, etc.
- Optimize the control of the weight of each animal and the feeding based on the controls.
- Reduce risks and mortality