AI startup tunes in to tomatoes with fresh funding: 'The potential is huge'

Sonicflora announced Thursday that it has secured €240k in a funding round led by Swedish state-owned VC Almi Invest. LiU Invest (the VC arm of Linköping University) and several business angels also participated in the round.

Impact Loop spoke to CEO Robin Jansson, who founded the company together with Henrik Niss and Oscar Davidsson, to find out more about their AI solution that aims to help growers understand what their tomatoes are saying about their health.

Until recently, Jansson worked as a freelance designer, but now he is fully focused on tomato tech.

“Sonicflora is about more than technology. It’s about giving growers a deeper understanding of their plants than ever before,” he says.

Linking AI to climate tech

It all began last year when Jansson joined Linköping’s lead incubator’s Boot Up programme, where entrepreneurs with a business idea receive support through lectures and workshops to develop early-stage ideas.

"I actually had another idea, but I set it aside when I met Oscar [Davidsson] in the same programme. He had this idea together with Henrik [Niss], and we decided to build on it together," says Jansson.

Niss’s mother runs a small farm outside Åtvidaberg, Jansson explains. Davidsson also has experience in farming. When the AI boom hit at the end of 2023, they started thinking about how to create something linking AI to climate tech. Around the same time, a study on plants’ sound signals was published, and they began investigating the subject in more detail.

How does the technology work?

"We have a hardware unit equipped with sensors that we place in a greenhouse," explains Jansson. "The units pick up ultrasound signals emitted by the plants and send them to a control unit connected to the cloud.

"There, the signals are interpreted and the grower gets information about what type of stress the plant is experiencing and where in the crop it occurs," he continues. "Each signal is shorter than 0.2 seconds and is nothing a human ear can perceive, but the data can be analysed."

Sonicflora has already developed a working prototype, but now aims to test its technology on a larger scale. Much of the development work focuses on collecting huge amounts of sound data, primarily from tomatoes, to train its AI algorithms.

Could reduce pesticide use

The goal is to give growers information about what their plants need to thrive, to minimise production losses and reduce the use of pesticides.

"80% of vascular plant species emit sounds," says Jansson. "We are currently training on tomato crops in Sweden. In parallel with entering the market, we will continue working with the next crop or ornamental plants. The potential is huge."

How specific can you be about a plant’s condition and its needs?

"That’s what we are working to find out. The research is new. Tomato plants emit up to 100 signals an hour, and after just a few signals, we can say with over 95% certainty what type of stress it is, such as water shortage. Grey mould is another thing we’ve started looking into, but we haven’t got as far with that. The model becomes more accurate the more data we collect," said Jansson.

Tomato plants emit up to 100 signals an hour

In addition to the new investment, Sonicflora previously received another €240k from the Swedish Board of Agriculture, in a project run in collaboration with researchers from the Swedish University of Agricultural Sciences (SLU), Agtech Sweden, and Linköping University.

The company will now use the latest funding to commercialise the technology and build the company.

"We will also continue developing the model to find more applications and build our own small test facility," says Jansson.