How AI is teaching students farming of the future

Microsoft has provided funding for a CSIRO pilot program that provides students with experience using AI to tackle agricultural challenges.

The Microsoft FarmBeats for Students initiative allowed students in Years 9 and 10 the opportunity to participate in a hands-on AI sustainable learning experience, applying smart farming techniques to food production.

Participating students and teachers received lesson plans aligned to the Australian Curriculum, tools and technology to explore how big data, AI, machine learning and Internet of Things technologies apply to real-world agricultural challenges through inquiry-based learning.

Additionally, teachers participating in the Australian FarmBeats for Students initiative received specialised professional development, helping them to use the technology and support their students. Schools received FarmBeats kits including a Raspberry Pi device as well as soil moisture, light, ambient temperature and humidity sensors to be used in student projects.

Using the INDRA platform, students were able to gain access to historical data from the local area to examine the changes that have taken place in the climate over time and to consider how those changes impact current and future growing conditions.

INDRA is a climate and hazard risk analytics engine designed to trawl big data collections and provide insights about challenges facing selected geographical areas. It forms part of the foundation for the Climate Services for Agriculture (CSA) digital platform, which was developed by CSIRO and the Bureau of Meteorology to deliver historical and predicted climate information at a 5 km² resolution.

Information from CSA about past and predicted rainfall, temperature, heat and frost risk, and evapo-transpiration is presented to farmers through a dashboard.

During the inquiry phase of the program, some students used the CSA prototype to plan their own investigation using the Microsoft FarmBeats kit. Students used CSA to understand how much rainfall their area had received over 10 years, taking averages of those data points and comparing it to the rainfall needs of plants they might grow for their investigation. Students also had access to Microsoft Lobe — a no-code machine learning application that students and teachers were able to train for various applications such as counting number of insects, recognising insect types and identifying ripe fruits.

Armed with all the resulting insights, students were encouraged to consider how to adapt growing practices for optimum yield.

“The teachers selected for the FarmBeats for Students initiative demonstrated passion for teaching AI. Many reported that the inquiry nature of the program allowed them to extend their students’ STEM skills, critical thinking and creativity. It was great to see so many different schools and teams participating in the pilot, which reached 397 students from 18 schools, including year 10 all-female digital technology students from St Margaret Mary’s College in Townsville,” said Ruth Carr, CSIRO Director Education and Outreach.

Image credit: Microsoft