AI-powered plant breeding start-up Avalo has secured US$11 million in funding to reduce emissions in the sugarcane supply chain. It is also partnering with Coca-Cola to develop low-nitrogen, low-water sugarcane for the beverage giant.

The North Carolina, US-based company will use the cash injection to advance its rapid evolution platform to commercialize resilient, low-input crops. Initially, the focus will be on cotton and sugarcane varieties that require less nitrogen fertilizer and water to thrive.

Each ton of sugarcane processed releases 26.5 kg CO2 equivalent into the atmosphere, resulting in 241 kg of CO2 equivalent for each ton of sugar produced.

Nicola Tongue, associate director at Coca-Cola Europacific Partners (CCEP) Ventures, says the partnership with Avalo can help the company reduce its scope three emissions, as ingredients represent one of the “hardest to abate” emissions areas in its value chain.

Meanwhile, a focus on evolved sugar varieties that use less water and synthetic fertilizer can “future-proof” the industry for farmers and producers, says Avalo’s chief marketing officer Nick Schwanz. It also prevents further damage to critically endangered ecosystems.

While introducing a new sugarcane variety usually takes over 12 years, he says Avalo could shorten the process to “five to six years.”

Germin8 Ventures and Alexandria Venture Investments co-led the funding round, with CCEP, Trust Ventures, Trailhead Capital, and angel investor Will Canine also participating. Existing investors AtOne Ventures, Better Ventures, SOSV, and Climate Capital also joined the round.

AI in agriculture

Avalo uses AI-powered Gene Discovery by Informationless Perturbation technology to identify genes responsible for key traits in crops, says CEO Brendan Collins.

He believes AI is set to have a “meaningful” impact across agriculture, including the large language models and computer vision for precision agriculture.

However, he adds that for crops besides maize and soy, there is not much historical data in genetics available to support these AI models substantially. “And that’s why we think our version of AI and machine learning is extremely valuable as we can make the most of the sparse and limited data that there is.”

According to co-founder Dr. Mariano Alvarez, Avalo’s model also speeds up traditional breeding practices. Conventional methods like genome-wide association studies identify the actual cause of a trait only 15-20% of the time, while Avalo’s model is “greater than 90% accurate,” he notes.

Meanwhile, scientists in China are also exploring AI and big data to overcome the limitations posed by traditional crop breeding methods.

Looking ahead

Collins says the CCEP partnership is a multiyear project, and Avalo aims to make it faster and less expensive.

By reducing the inputs that have to go into the field, the company also plans to enhance sustainability and farmer profitability, and supply farmers with seeds at lower prices.

Avalo aims to have its first machine learning models trained this year to start making recommendations at the end of this year or at the beginning of next year, says Alvarez.

“We really want to make an impact in the supply chain in the next three to five years.”