A mobile-based Internet of Things (IoT) system is helping farmers support one of humanity’s biggest food chain benefactors: the humble - but absolutely vital - bee.
Using a combination of 2G-based SIM cards deployed on special boxes in fields, suppliers of produce to major high street chains, including Tesco and Marks and Spencer, as well as brands such as Innocent Drinks, can take remedial action to boost the numbers of ‘pollinators’ (a bug that conveys pollen to a plant and allows fertilization).
As a result, CEO and founder of UK-headquartered company AgriSound, Casey Woodward, says users can plan for a factor of 10 increased pollinator visits for their next growing season.
That matters for their individual farm’s bottom line, as sub-optimal pollination has been shown to cost growers in the region of £5,000 per hectare.
But as the global market value linked to pollinators is estimated by Friends of the Earth as being worth as much as $557bn each year (in the UK alone, the value of the crops they help grow is worth £691m a year) having humans do the work could cost the food sector at least £1.8bn a year.
Helping bees and the other insects to do this ‘job’ also ultimately may benefit the planet, as they pollinate 70 of the 100 crops that provide 90% of the world’s food - which is why AgriSound sees such value in what it calls ‘an Alexa for bees.’
Founded in 2020 by Woodward with 10 staff, AgriSound is dedicated to using IoT and wireless communications in agritech.
It accelerated this move by purchasing a specialist in remote beehive monitoring tool, Arnia, which had been operating in the market since 2009 across 14 countries. The combined firm now has customers in 25 countries.
Arnia’s product uses a specialized MTM (machine-to-machine) SIM card from cellular connectivity firm Pod Group. Since 2012, Pod is now part of a German networking company called Giesecke+Devrient, which claims to have launched the first commercial SIM card in 1991. Woodward says:
Everything now can be smart, everything can send data and can send information, but you need connectivity to make that happen because the data has to be sent to a back end, and therefore you need connectivity.
I'm a scientist by background but have spent a lot of time working in the agriculture industry, seeing people using advanced sound analysis to listen out for the welfare of different types of farmyard animals, like the way their pigs were coughing.
I got to thinking, could we use those sorts of technologies for monitoring bees? We know they are dying out for several different reasons, mainly driven by agricultural intensification, but we can't take action unless we know where to apply a possible solution.
So, we’ve built a suite of digital signal processing techniques, in combination with machine learning, and our vision is to have the world's largest insect biodiversity database, which we can share with governments, multinational agrochemical companies or bee suppliers as needed.
Polly - a digital beekeeper
That tech is embodied in ‘Polly,’ the company’s in-field ‘apiculture’ (beekeeping) monitor.
Powered by onboarded solar panels, a Polly can sit on a pole or be embedded on the ground. Best practice indicates that between two and four devices per hectare delivers optimal coverage.
Inside each Polly is a microphone that listens for insect sounds, as well as sensors for temperature, humidity, and light.
The machine is sensitive enough to detect if individual, or swarms, of honeybee, bumblebees, wasps or other flying insects are passing.
Once operational, the devices generate millions of datapoints per day.
These are transmitted in the form of messages like, ‘There have been 10 pollinators the last hour in this part of your field,’ along with contextual environmental data.
This data is transmitted via a high reliability 2G multi-network which searches for the strongest local carrier signal to work on to a central control to understand hive (or ‘colony’) status in near real-time.
A key data point is wingbeat frequency, which can tell the software a lot about the viability of a passing flying insect.
This data is then analyzed and predictive ‘bio-acoustic’ algorithms are applied to support beekeepers in a range of pollinator husbandry tasks, like when to carry out hive inspections.
Analysis is delivered to customers via a web portal, where they can see various metrics on pollinator activity, such as ‘heat maps’ of activity which indicate where bees might need some assistance, or where placing a beehive might produce greater pollination or higher potential for a bigger honey harvest.
That ‘one plane of glass’ management is what attracted him to the network equipment supplier. He adds:
The main reason we like the product is the dashboard; we can turn stuff off, we can turn stuff on, we don't have to worry about if we need to manually swap stuff over to a different network, and so on.
And as we have devices out in a lot of different countries it’s great to be able to manage all that from one central place, as that makes our customer support work much easier and so allows us to be much more agile. For customers, doing this with technology is also much better than the only other alternative, which is someone standing in the field with a clipboard.
Another benefit is that in rural areas, at the edge of fields, or in the wildflower meadows pollinators congregate in, being able to easily acquire the most reliable network signal also guarantees data collection, he adds.
Next steps for this combination of IoT and 2G to help keep the humble bee do its valuable job for clients, he says, will be adding the ability to differentiate between different insect species, and so give customers the ability to tell the difference between all the various pollinators helping them out.