Digital technology adoption varies tremendously across the globe. Asia leads the way in terms of digitalization, with daily activities, such as paying for public transportation and ordering food in restaurants now managed through smartphone applications. E-commerce in the region is booming.
Globally, the poultry and egg sectors, so rich in valuable data, have been slow in embracing digitalization. Doing so could help with attracting workers and keeping them. Where flock management is concerned, it should come as no surprise that younger farm workers expect the wide range of smart applications to assist them that they have grown accustomed to in other areas of their lives.
Asian example
In egg production, checking daily mortality is a difficult task in the high-rise cage systems used in Asia, further complicated by worker safety regulations requiring use of harnesses and other safety devices above a certain height.
Robots can now detect dead birds in cage systems following supervised machine learning, one of the latest programming techniques derived from artificial intelligence.
The initial training stage consists of taking photographs of hens in their environment and labeling them to clearly identify which bird is alive and which bird is dead. While this machine learning step previously needed up to 10,000 pictures for training and several months of work to label each shot, newer models require under 2,000 pictures, while labeling has become semi-automated, using other machine learning-derived algorithms to reduce the process to under two weeks.
Once the training has been validated and tested with additional pictures, the robot is ready for incredibly accurate mortality detection and, unlike humans performing this task, is never distracted during the inspection.
The robot, equipped with as many cameras as needed to thoroughly inspect each tier in a cage system, moves along the aisle at a speed of 12 meters per minute, needing about eight minutes to complete the full length of the barn. With each cage identified through a QR code, the robot identifies the cages that will require attention, relaying its observations to a display panel, often mimicking the type used in video games favored by younger generations.
While inspecting each cage, the robot is also gathering valuable information on temperature, humidity, ammonia level and wind speed, all parameters providing a more granular vision of the environmental conditions in the house.
Although Chinese researchers are now dreaming of a smarter robot able to autonomously remove dead birds from cages, the intervention of farm workers remains essential to identify cause of death and to inspect the remaining birds for any signs of disease.
Broader benefits to come
Robots could also be trained to count birds in cages, gauge feather coverage levels and even detect early signs of diseases. Any visual observation made by a skilled farm manager can be taught, through machine learning, to robots equipped with cameras.
Embracing new technology tools will not only attract younger generations of farm workers to our sector but also empower them to focus on added-value activities that will contribute to higher productivity and improved welfare. Collaborative intelligence is the way of the future, with people and robots working together.
Layer-house-robot
A Chinese smart robot, trained trough machine learning, can detect bird mortality and continuously monitor flocks. Dr Vincent Guyonnet
SIDEBAR
Global egg production still going strong