Many technologies already exist in farms to monitor animal activity and health. In cows in particular, accelerometers in the collars, or pedometers attached to the legs, measure activity and detect periods of heat, lameness, rumination time, etc. These tools are generally very useful to breeders, allowing them to save time and improve the precision of herd management, especially when the latter are growing.
Now a new generation of sensors is coming onto the market, with which animals no longer wear collars and bracelets. Like, for example, the AIHERD solution presented by the Nantes-based company of the same name at the international livestock show Space, in Rennes, in September 2022.
This innovation was rewarded with an Innov’Space innovation trophy. Using cameras positioned every fifteen to twenty meters in the building (€430 each), the cows are constantly observed. Each is identified by the stains on her dress forming a kind of QR code.
Time budget for each cow
« The cameras associated with the analysis of images by artificial intelligence record the time budget of each cow, explains Édouard Asselin, commercial director of AIHERD. The system measures the time spent eating, drinking, walking, standing or lying down, as well as interactions with other animals. The behavior makes it possible to identify if the cow is in heat, if she has a health problem, or if a calving has started. »
If the eye of the breeder remains essential, the advantage of these tools is to ensure permanent monitoring of each individual, and to identify anomalies early. Once alerted, the breeder can check on his own, and quickly implement first aid before a degradation can lead to heavier treatments or loss of production.
Regarding the management of reproduction, the system is able to detect so-called silent heat, that is to say, the expression of which is very discreet or even absent. It would be in this area more efficient than the existing tools on the market.
Not suitable for grazing systems
However, AIHERD has a drawback: the animals must spend at least half the time inside the building under the eye of the cameras. The tool is therefore not suitable for very grazing-oriented systems where animals sleep outside.
This new monitoring solution promises for the future other applications such as the measurement of the body condition score of animals (BEC). “Today we are only able to see if this NEC is increasing or decreasing, says Edouard Asselin. We are also working on the observation of the respiratory rate in order to detect any thermal or pathological stress. »
Ro’Peek: the aerial robot that monitors the atmosphere
Moving on a rail, the Ro’Peek robot assists the breeder in monitoring the buildings. © Copeeks
The young Breton company Copeeks based in Lannion (Côtes d’Armor) also received an Innov’Space, for its aerial robot for the supervision of sheltered farms Ro’Peek. Weighing three kilos, moving on a rail of one hundred meters maximum, it is equipped with a 360° camera and a battery of sensors (temperature, humidity, CO2, ammonia, fine particle rate, etc. ). It collects both images and mood settings. This monitoring makes it possible to detect any heterogeneity in the spatial distribution of the animals, likely to reflect a deteriorated environment (litter, ventilation, lack of water or food).
“The breeder goes through his building once or twice a day to identify anomalies or incidents, says Gwenaël Le Lay, founder of Copeeks. With this tool, he saves time and targets problem areas more easily. »
The pork cooperative Evel’Up, a partner of Copeeks for several years, offers its breeders a diagnosis of the living spaces of pigs. « Without the influence of human presence, it is possible to identify, for example, the feverish behavior of piglets several days before the breeder detects the signs of influenza », says veterinarian Guillaume Jousset.
Pig bites caught in the act
Identifying tail-biting phenomena aims to better understand and anticipate them. © Nathalie Tiers
Based in Rennes (Ille-et-Vilaine), the young company Dilepix presented to Space a prototype for the automatic detection of tail biting (tail biting) in pigs, using video analysis and artificial intelligence. Called “Tail” (“tail” in French), this tool combines a removable camera placed in the collective boxes of post-weaning or fattening pigs, with an image analysis application using artificial intelligence algorithms. The behavior of pigs is therefore not altered by human presence. Tail is not intended to be used by breeders themselves. Dilepix is currently developing it with the pharmaceutical company Boehringer Ingelheim with the aim of better understanding and anticipating tail-biting factors (boredom, temperature, diet, etc.).
Dilepix is also developing with the Inrae Institute the Sowell tool for measuring the activity of a group of pigs. She is working on the individual identification of each pig using her tag.