Fully digital ranch at ILRI in Kenya to transform livestock research

At first glance, the ranch operated by the International Livestock Research Institute at the Kapiti research station and wildlife conservancy in Kenya looks similar to other ranches seen across eastern Africa. The only sign of an exciting new project are the electronic collars worn by some of the cattle. They enable scientists to create accurate digital representations of the cattle in Kapiti so that they can understand their interaction with the environment. These so-called digital twins are a powerful new way of studying agriculture and environment. At Kapiti, the digital twins are being used for research into issues such as rangeland management and climate-resistant cattle.

The Bluetooth collars worn by the cattle are provided by the South Korean technology company Bodit, in a partnership under the CGIAR Initiative on Digital Innovation. Each second, the collar streams dozens of data parameters on the cow, which are processed by machine learning to track their movement, behavior and health.

A second sensor is deployed into the cow’s rumen where it remains for its lifetime. The bolus sensor provided by Austrian company Smaxtech under the CGIAR Sustainable Animal Productivity (SAPLING) Initiative tracks data such as internal body temperature, water intake and other signs of animal health and activity.

Finally, a full 3D render of the ranch, soil map and automated weather systems provide a virtual model of pasture availability.

All the systems are now online, with the goal of integrating all their data in a single system by the end of 2024. Taken together, ILRI now possesses virtual representations of complex natural systems: to monitor animal health, track grassland management, and phenotyping for animal breeding programmes.

“The question of optimum management of rangelands is a national-scale problem, especially for eastern Africa,” said Ram Dhulipala, senior scientist at ILRI. Digital twins for agriculture have shown promise in a few cases but, according to a 2022 study, research is needed into systems for grassland management that could be adapted into different use cases. The ranch at ILRI will act as a test-bed and showcase for systems that could be adopted by larger commercial ranches with a focus on scalability and reliability of the digital infrastructure.

Another challenge is in animal breeding: conventional practice requires researchers to be in the field observing animals directly. The ILRI digital twin will observe how cows with different genetic backgrounds associated with heat resistance fare, to see if panting or signs of other respiratory distress can be a reliable indicator of heat sensitivity and as a test case for real-time digital phenotyping. “The digital twin for phenotyping will be an effective tool which caters to livestock researcher communities improving the health and welfare of cattle and ensuring that farmers have access to the most suitable breeds for their needs”, said Dhulipala.