AI-powered soil analysis software wins Imperial’s undergraduate competition

Credit: Brendan Foster Photography

Team AmmonAI are on a mission to create ‘greener fields and smarter yields’ with their tool that guides smart farming decisions.

AmmonAI, a software for farmers to measure soil quality and make smart farming decisions, has been awarded £7,000 at Imperial’s Faculty of Natural Sciences Make-A-Difference (FoNS-MAD) competition. 

FoNS-MAD is Imperial’s only competition that offers undergraduate students a funded lab placement to develop a startup idea into a proof-of-concept. It aims to equip students with the skills, knowledge and resources to develop low-cost technologies for the benefit of society. 

Smart farming 

Team AmmonAI said that fertiliser has 'seen a rapid increase since the 1950s.' Fertilisers usually contain nitrogen, which can lead to algae blooms and ecosystem damage. The team want to ‘empower farmers’ to make informed decisions to avoid over-fertilising their fields, while still maintaining efficient yields.  

The software, AmmonAI, would use data input from a soil probe and help farmers predict future nitrate levels in soil, considering factors like weather. The lab placement saw the team grow spinach in different conditions and successfully predict soil Nitrate-Nitrogen levels, with 93% accuracy. With the winning money, the team hope to train the model with a larger range of conditions.  

Team AmmonAI is made up of Ritsuki Takesako in the Department of Chemistry, Sarkis Paul Shadarevian in the Department of Mechanical Engineering, Stephen Heslip and Paolo De Biase both in the Department of Mathematics. 

The panel of expert judges, made up of Professor Lord Robert Winston, Professor Sir John Pendry, Dr Ruth Allan, Dr Allan Samuel and Dominique Kleyn, described AmmonAI as having the ‘most potential for commercialisation’ as well as high levels of ‘academic rigour.’ 

Opening the award ceremony, Imperial’s President Professor Hugh Brady said: "Since joining Imperial, I’ve been truly impressed with our students' passion for innovation and their curiosity to tackle the major challenges facing society and our planet.

"It’s clear that our students, and indeed our staff, join Imperial because they want to make the world a better place, and competitions like this are vital to help turn these bright ideas into a reality."

On winning the award, the AmmonAI team said: "We are so pleased to be the winners of this year's competition. Overuse of fertilisers is not just a national issue, but rather a global one. There is lots of work to come next, and we will take on board all the useful feedback to continue to develop our idea. 

"We would urge anyone thinking of applying to this competition to just go for it. Not just for creating a startup, but for the networking, new skills and working outside your comfort zone too."

Lab placements  

Three other finalist teams also pitched at the ceremony.

Arsenic removal – Ecoflow 

The Ecoflow aims to use fruit peel to remove arsenic from underground well-water in Asia. The team powdered and treated common Asian fruits with different chemicals to identify which combination could remove arsenic most efficiently. Certain fruit peels – mainly banana – removed arsenic by at least 90% in the trials. Students Anna Xia (Department of Chemistry), Shan Ding (Department of Life Sciences), Hongzhi Wu and Kazu Osuka (Department of Bioengineering) make up team Ecoflow.

Biodegradable ‘plastic’ film – Licaze

Licaze aims to use waste products to produce sustainable materials. Over their placement, the team extracted lignin from weeds and chitosan from crab shells and combined the two to make a bioplastic film. The film was a stronger and water-resistant alternative to standard packaging. Students Yong Ji, Chan Li, Hongyang Zhu and Ruchita Duva from the Department of Life Sciences make up the Licaze team.

Water monitoring – Qualboat 

Qualboat is a low-cost and user-friendly water quality surveying tool. The proof-of-concept prototype could plot its own route across a body of water to collect data on pH, turbidity and total dissolved solids in real time via a simple app. It demonstrated a three-hour running time with 92% accuracy. It was designed by Charlie Campbell (Department of Computing), Aidan Randall (Department of Physics), Julia Gong Pinho (Department of Bioengineering) and Sixian Zheng (Department of Mechanical Engineering).

Past success 

In the ten years since its inception, FoNS-MAD has fostered great ideas from its students. Previous winners include Multus Biotechnology, a cultivated meat startup that raised £7.9m in funding after winning FoNS-MAD in 2019, as well as Matoha, a plastic and textile identification device which won the competition in 2017.