Scentian Bio raises $7m to take its AI and insect-powered ‘digital nose’ to market
“Imagine if you could detect a disease on your breath simply by blowing onto your phone,” Dr Andrew Kralicek says.
More than two decades ago, it struck the Auckland academic that an incredibly sensitive detection system already exists in nature.
It was already known that flies could be trained as “sniffer dogs”.
A fly’s antennae can be used to detect cancer cell lines, as well as explosive and drug compound precursors," Kralicek says.
“Bees, which I love to call the dog of the insect world, because they can be easily trained, can be used to detect Covid, explosives and pests in the field,” he says.
Kralicek estimated it would take three years to produce a biosensor chip that could replicate the receptor proteins that insects use for their super-sense of smell.
In the event, it took 20 years. But by 2023, he had a prototype that was good enough to attract $3.5 million in funding from the Bill and Melinda Gates Foundation, which saw potential for the technology to detect malaria.
Kralicek said it would take another two years, from that point, to get the technology to the market.
This time, he was right, or as close as you get to it in the timelines of early-stage companies.
Scentian Bio, the start-up that the veteran Institute of Plant and Food scientist founded with one-time Fonterra business development manager Jonathan Good in 2020, is now just months away from commercialisation.
To get it over the line, the start-up has just raised $7m in an over-subscribed “pre-Series A” round led by Icehouse Ventures, which now holds a 19% stake after also supporting an earlier seed round.
The money will be used to scale up manufacturing and create sales channels.
The raise was also supported by new investors Cultivate Ventures and New Zealand Growth Capital Partners, through the Crown agency’s Aspire fund.
Existing investors Toyota Ventures, Booster, DYDX Capital and the Tindall family’s K1W1 also chipped in more funds.
Kralicek’s employer for 19 years of his research, Plant and Food (now part of the New Zealand Institute for Bioeconomy Science), holds an 18% stake.
Kralicek’s goal remains a “digital nose” in every phone, with taste and smell receptors rounding out the five sensors (given your iPhone or Android can already see, hear and sense touch).
But the immediate future is a corporate play, detecting food and beverage quality.
“We’ve got pilots with seven leading global food companies,” Good tells the Herald.
Most are based in US.
These include: “Food and flavour firms and global leaders in spices and condiments, as well as beverages.”
For food manufacturers and ingredient suppliers, quality issues that show up late can trigger waste, rework, disputes and reputational risk, Good says.
The idea is that ingredients can be tested as they come in the door, heading off problems later in the process. It positions Scentian as “the recall killer”.
Kralicek says every organism produces volatile organic compounds (VOCs) – or what most of us would simply call smells, if we could detect them all with our human noses.
There are more than 2000 volatiles on your breath, he says.
There are already sensors that can detect the bad odours that indicate food spoilage and more.
“But they’re either too slow so can’t provide an answer in real time, or not accurate enough because they don’t have the sensitivity or selectivity required, and if they are accurate, they’re hard to access because they’re large machines housed in centralised labs, needing an expert with a PhD to not only run the machine, but also to interpret the data.”
Scentian Bio’s sensor on a chip, on the other hand, can be held with two fingers.
And you don’t need a doctorate to interpret the results. They’re processed in minutes by artificial intelligence (AI).
Kralicek says his eureka moment came when he realised how to replicate insect protein receptors, honed over 400 million years of evolution, on a chip.
He says the accuracy is down to “parts per quadrillion. That’s equivalent to one drop of water in 20,000 Olympic-sized swimming pools. So when we attach insect olfactory receptors to sensors, they are 1000 times more sensitive than a dog’s nose and nose, and a million times more sensitive than the human nose”.
But while the system was incredibly sensitive to volatile organic compounds, it also had to be taught what each combination of VOCs actually meant. That’s where AI training came in. Scentian’s system can now detect and interpret smells that no insect would normally encounter in its lifetime.
Scentian is initially focused on food, but the underlying sensing technology also has applications across human and environmental health. Technologies with that breadth of application are rare, and it will be exciting to see how this ‘digital nose’ develops,” Icehouse Ventures principal Bex Gidall says.
Its initial food and beverage quality monitoring market is worth around US$8 billion a year.
Good says: “There are lots of other verticals, lots of opportunities.”
Gidall says they include pest detection, disease diagnosis, explosive detection and air quality monitoring.
Could we all be wearing lapels with Scentian Bio sensors built-in, one day?
“Or you could just point your phone in the direction of what you want to analyse,” Kralicek.
That could be some street food in a developing country, or some avocados as you browser the supermarket and wonder if they’re ripe.
“Or it could be your smart watch tracking metabolites in your sweat,” Kralicek says. (Metabolites are molecules produced when food, drugs or bodily tissues break down, and can signal if you’re healthy.)
Today, Scenitan has a liquid sensor. Samples are applied to its device’s 2048 membranes via syringe - the usual form for the food and beverage industry. But it already has a prototype gas sensor, which will allow a much wider variety of uses, for a much wider audience.
Chris Keall is an Auckland-based member of the Herald’s business team. He joined the Herald in 2018 and is the technology editor and a senior business writer.