Technology for More Sustainable Agriculture

Winner of the Junior Academy Challenge – Greetopia Team

“The Green Redesign: Improving Energy Efficiency in Cities Using 5G and IoT” – Fall 2022

Sponsored by Ericsson

Team Members: Jiho L. (Team Lead) (Republic of Korea), Ansh T. (India), Riya K. (India), Arshroop S. (India), Aman A. (India), Rawnaq A. (Oman)

Mentor: Olusola Ladokun (Nigeria)

---

Among the 85 teams that embarked on the Green Redesign Challenge in Fall 2022, one team stood out.

By improving irrigation and reducing the waste of resources, Team Greetopia aims to bring innovation to agriculture. They were named the winning project in the challenge.

Based on research and stories shared by numerous farmers, the team identified key issues, illustrated by stark figures: 2 quadrillion gallons of agricultural water are wasted annually^[1], and agriculture uses 2.2 quadrillion KJ of energy every year^[2], about 5% of world energy demand. They also found that 1.2 billion tons of food^[3]– enough to feed the entire US population for 3.64 years– is wasted, never leaving the farm.

Working online, across time zones, created initial difficulties– but the team members found ways of collaborating productively.

“I’m grateful to have experienced the diverse cultures coming together for the betterment of this human society,” says Aman. “Time zones were a hurdle in the smooth performance of the team, but we managed it by distributing the work evenly to be performed by individuals at their time of comfort, alongside holding alternate team meetings at common times.”

Challenges in Agriculture

The Organisation for Economic Co-operation and Development (OECD) has identified lack of innovation as one of the biggest challenges in agriculture. In a world facing multiple challenges such as depleted natural resources, climate change, and pollution, developing more efficient agricultural systems is crucial for our survival.

Through online brainstorming, the team concluded that emerging technologies like the Internet of Things (IoT) and Artificial Intelligence (AI) can address some of these issues and can help farmers who often lack the skill set to optimize agricultural methods. Overuse of fertilizer, for example, can pollute waterways, burn crops, deplete the soil of minerals and increase air pollution.

“It was an amazing experience to have connected with like-minded individuals and research world-changing approaches!” says Riya.

Greetopia team members decided to develop a web application that would tackle the excessive use of non-renewable resources.

“I got to research and learn more about the important crises around the world, such as concrete pollution, irrigation, etc.,” says Arshroop. “The constant communication within the team allowed us to keep up with the information and learn a lot of valuable lessons through the program.”

Utilizing the Internet of Things

In particular, the students opted to use the Internet of Things (IoT) to increase efficiency in irrigation and modernize farming practices that have remained unchanged for centuries.

“As team lead, I worked on planning our work, informing each member of what they have to do until certain target dates, creating documents to make the process efficient, and of course, conducting research about our research topic and filling in the milestones,” explains Team Lead Jiho.

The team devised “Kanad”, a farming application that delivers four main functions. First, it senses soil moisture levels by using a machine-learning component called Long Short-Term Memory Network (LSTM), which gets more accurate with use. Second, farmers can enter information on the Nitrogen, Phosphorus and Potassium (NPK) content in the soil, (also analyzed by sensors) to identify the most optimally adapting crops for these soil characteristics and locations.

Utilizing Machine-Learning

Machine-learning can also recommend the optimal amount of fertilizer based on the same NPK levels in the field. Finally, farmers can enter images of their crop into the web-based application, which will use the Convolutional Neural Network deep learning system to identify potential crop diseases with an accuracy of 95.25%.

Arriving at this successful solution did not prove easy: the team decided to change direction halfway through the Challenge.

“I’m glad that the team members understood the sudden change in our projected target in the middle of the program, which led to a successful ending,” says Ansh. In the end, perseverance paid off and the outcome amply justified the effort involved.

“Despite the obstacles we encountered, we made it till the end,” says Rawnaq. “We did not just complete the challenge but the result was awesome.”

---

[1] Source: UN Food and Agriculture Organisation
[2] Source: US Department of Agriculture
[3] Source: World Wildlife Fund (WWF-UK)