Info Patronale · No. 332 · May 2025

Bionomous: automation for biological research

How did the idea for Bionomous come about?

English version

Interview with Frank Bonnet, founder of Bionomous

The idea for Bionomous emerged during my PhD at EPFL, where I developed a robotic fish to study the social behaviour of fish in groups. It was an original European project aimed at integrating robots into groups of fish to observe their social behaviour. My job was to design these robots, while my biologist colleagues in Paris carried out behavioural experiments with real fish.

Every time I joined them for the experiments, I noticed that they spent a huge amount of time sorting fish eggs by hand rather than using the robots I was working on. Initially as a joke, I suggested the idea of a robot dedicated to sorting eggs automatically. Then, attending conferences on zebrafish, I realised that this need was widespread beyond my team. That is how the desire to develop a more widely applicable automated solution was born.

What exactly is the solution offered by Bionomous?

Our technology combines robotics, fluidics, optics and artificial intelligence (AI). At the heart of our system is a patented method that automatically isolates biological entities (such as fish eggs) and presents them individually to a camera for automatic sorting based on image recognition.

Initially, our AI algorithms were developed exclusively in-house. Today, we allow users themselves to train the models according to their own specific needs. This makes our solution extremely adaptable to different applications and contexts.

What exactly is the role of artificial intelligence in your system?

AI is mainly used for image recognition. Each item to be sorted is presented in front of a camera, which captures an image. The image is then analysed by machine learning-trained algorithms capable of distinguishing, for example, fertilised from unfertilised eggs. As we collect more data from our users, our AI model becomes increasingly accurate and powerful.

Who are your current customers and what are your potential markets?

At present, around 80% of our customers come from academia and research laboratories. The remaining 20% are mainly from the pharmaceutical industry, including small companies that provide services to the large pharmaceutical groups.

We are currently looking to extend our applications beyond the strictly academic sector, in particular to aquaculture (for fish hatcheries) and agriculture (for seed sorting). We focus on items of around one millimetre, a segment still little exploited on today's market, which represents an excellent development opportunity for Bionomous.

How do you see automation evolving in the laboratories and industries concerned?

Automation is booming and is not limited to biology. In a scientific context, it responds to a growing need for standardised processes. It ensures consistent results, which is essential for the pharmaceutical industry in particular when testing and validating medical treatments.

An emerging trend is organoids, small biological structures that can reproduce human tissues such as the liver or brain to test treatments without directly using animals or humans. Current processes for handling these organoids are very expensive because they are highly manual and time-consuming. Automating these processes would considerably reduce costs and speed up research cycles.

What benefits do you gain from your collaboration with Fri Up?

Being part of the Fri Up network gives us the opportunity to join a dynamic ecosystem, making it easier to interact with other start-ups, venture capital investors and potential partners. Thanks to Fri Up, for example, we were able to make several key contacts to explore new markets and strategic partnerships.