The financing is a co-investment led by Obvious Ventures and Aqua-Spark, as well as Real Ventures, which previously invested in the company’s seed round.
The aquaculture industry supplies more than half the seafood consumed worldwide and is growing at 6% per year. However, the Canadian firm said the use of outdated technologies like hand counting of aquatic organisms results in major financial losses and waste. Fish farms can overspend by 20% in feed alone, and fish survival rates can be as low as 50%.
It claims its technology, XpertCount, on the contrary, combines artificial intelligence (AI), computer vision, and machine learning so that hatcheries, farms and research centers can precisely track and manage early-stage aquatic organisms such as shrimp larvae and live feed.
Valerie Robitaille, CEO and Co-Founder of XpertSea, which is based in Québec City with offices in Montreal, said the platform is already in use in over 100 facilities worldwide.
The $10m capital injection will go towards both market expansion and the company’s engineering capabilities, so as to provide even more applications and data for the industry:
“We want to build our team of data scientists. We use a lot of machine learning and AI so we want to double our engineering team,” she told FeedNavigator.
It was in 2011 that the team had the germ of the idea for XpertCount:
“We visited a lot of farms. We spent a year in the field. In 2015, we developed the current platform and we launched it in 2016. Our goal was to develop a solution that was suitable to most commercial scale aquaculture operations.
“We have applications for every segment from shellfish to marine fish to freshwater fish, even trout. We are mainly in the hatchery, nursery phase - the beginning of the value chain. However, in shrimp, we are in the entire value chain.
“We also work with many aquaculture research institutes that need to have accurate monitoring of fish survival and growth.”
One of the key elements of this IoT quality control platform is the inventory management:
“A lot of other techniques for estimating the total number of animals [at the hatchery stage] are not very accurate. Using our device allows managers to determine the number of organisms [in the tank] with 95% accuracy, so when they stock their tank they know how many animals they start with, and that really helps with optimizing the feeding regime.
“Managers can take measurements every day and follow the growth. If they see that they are starting to have a big differential in the size of the organisms, they can adjust the feeding regime accordingly. Sometimes the feed pellet size has to be matched with the size of the mouth of the shrimp, for example, so being able to have a good assessment of the size distribution of the organism can help.”
XpertCount pools different technologies – computer vision and optical measurements. “Each technology had its limitations but when combined together they are complimentary and they allow you get a lot more information about what is going on. It is very complex in these environments - there are other types of particulates in the water, and then there is the behavior of the animals [to contend with], they tend to cluster. That is why we have to use very advanced AI and machine learning techniques to be able to accommodate for all these cases.”
The investor perspective…
Mike Velings and Amy Novogratz, co-founders of Aqua Spark, said fish farmers cannot make accurate judgements, particularly about feed, if they do not know the number of organisms in the water or distribution size. “XpertSea is the first company with a successful product to focus on smaller organisms suitable for hatcheries, which is where the farming process begins. Their ability to solve production challenges in aquaculture is limitless and makes them a key value-add for our portfolio.”
Andrew Beebe, managing director at Obvious Ventures, said XpertSea is digitally upgrading the world of aquaculture in a way that will benefit all players from farmers to retailers and consumers.
“We typically look at companies tackling really big problems. This technology will help make farmed seafood safer, more affordable and more sustainable.”
The platform, he continued, will help reduce waste and make farmed fish production more efficient by ensuring the early process within hatcheries, across species, is more trackable.
Machine vision in this space was long overdue, he said
“We are very enthusiastic about the team.
“Out here in Silicon Valley and, around the world, in tech hubs, there is a lot interest in artificial intelligence. However, I think the gap being bridged here is the ability [of a company] to deeply listen to, work with and understand its customer’s needs and to be able to apply technology accordingly.
“We looked at hundreds, if not thousands of aquaculture focused companies. and, one of the things that got us excited us about XpertSea is that [the team] literally moved to Asia to get to know their customers, they became embedded, and they then designed technology to do a complex thing in a simple way – that is product management and product design at its best.
“Oftentimes, we say there is a technology in search of a problem - computer scientists can be on to something and have great insights, but they don’t necessarily know how to connect that to a customer. So this is a story about artificial intelligence being embedded within an industry that has been slow to change and has been doing things in a similar fashion for a very long time, but it also a story about a company that is very in tune with its customers’ needs.”
The technology can also help feed providers understand the efficacy of their feeds within hatcheries and farms, he added.