Editor’s Note: AgFunder is an investor in The Yield.
“We put electronics into open bays in one of the world’s roughest marine environments with an objective to protect food safety; it doesn’t get harder than that.”
Ros Harvey, CEO of ag sensor and IoT startup The Yield, is talking about the Australian startup’s oyster-focused aquaculture business line. Founded in 2014 with a mission to “feed the world without wrecking the planet,” The Yield purposefully focused on a specific business problem, in a relatively small market segment, to perfect its technology stack before expanding. The startup now offers an end-to-end microclimate sensing and predictions service for land-based crops as well as aquaculture.
The specific business problem faced by oyster farmers relates to the harvest restrictions placed on them by regulators. Food safety regulators use rainfall as a way to determine when to open and close oyster harvesting because rain causes run-off from fields and surrounding land that could include dangerous contaminants, which will be filtered by the oysters.
Measuring the salinity of the water is a much more accurate way of determining run-off and the presence of these contaminants. By deploying microclimate sensors measuring salinity, The Yield can reduce unnecessary closures by up to 30% by providing this information to both growers and local regulators. This translates into an extra four weeks a year of harvesting time, which is not insignificant for farmers, according to Harvey.
The technology soon progressed to use machine learning and artificial intelligence to offer oyster farmers salinity predictions used by the regulators to open and close harvest with 95% accuracy three days ahead of time.
The end-to-end service, which involves the deployment and management of the sensors as well as the data analytics and user interface, also enables oyster farmers to manage labor scheduling, disease risk, harvest records, and consignment reporting. For example, growers are required to keep track of water temperature and salinity when shipping out oysters as part of food safety standards.
The Yield sells the oyster product to the food authorities in the states of New South Wales and Tasmania, representing over 60% of Australia’s oyster industry, according to Harvey.
“We really learned to do this in some of the most difficult conditions in the world, and to some of the highest standards out there as it relates to food safety, so it was a great place to launch our next product from with all we had learned,” she said. “It’s also a great idea to start small and localized to iron out issues first before you scale.”
Onwards to Agriculture
Sensing+ for Agriculture, which the company launched last month, uses the same concepts of microclimate sensing and machine learning-based predictions, to offer an end-to-end solution for intensively irrigated crop growers.
“Almost every grower we talk to has at least one unused agtech solution sitting in the shed, because they are rarely installed or supported properly, so it’s core to our offering that we take care of the whole package – from installing, supporting and maintaining the sensors, to delivering predictions and insights via mobile apps and on the desktop; all hardware and software is supplied as an ecosystem,” says Harvey.
The Yield built the product alongside growers in Australia over two years to ensure it was solving the specific challenges they face around the five main business decisions they need to make each year: planting, irrigating, feeding, protecting, and harvesting.
Sensing+ now offers farmers informed and intuitive decision support around these five activities by measuring the 12 factors that drive most agricultural models: temperature, rainfall, leaf wetness, humidity, soil temperature and soil moisture, wind speed and direction, photosynthetic active radiation, total solar radiation and air pressure.
Using irrigation decisions as an example, The Yield uses the microclimate data it collects to create a water balance for farmers — including the amount of water lost to evapotranspiration and the growth curve of certain crops –and to predict the soil moisture at any given time. This enables the farmer to adjust their irrigation schedule and optimise the use of costly water resources.
“We’re the only business making 7-day predictions for soil moisture – and microclimate growing conditions generally – wherever we have hardware,” says Harvey.
This information is presented to growers in an app. Growers use the app to learn, act and record activity that generates an auditable train which Harvey describes as a “news feed for each field and crop.”
Timely Decision Support
For spraying decisions, the Sensing+ app gives growers the microclimate insights they need to decide the best time to spray based on the 12 factors such as humidity and wind, which might make agrichemicals less effective with the risk of washing off the plants and getting into the environment. Sensing+ also helps growers track when spraying actions occur — which is particularly relevant for large-scale operations that sometimes struggle to keep up with all the activity on their land. And it does all the relevant spray calculations, records things like withholding periods, and gives a real-time inventory based on the amount of chemicals used. Each spray action creates a date and weather stamped record as part of the field and crop record.
Growers can also choose to receive custom notifications in the app, where they input the conditions they want to know about, and Sensing+ sends them an alert when these have been reached. One example is wind speed and direction, as growers will rarely spray when the wind reaches higher than 20 km per hour, so they can set to receive an alert when the wind is 18 km per hour and then turn off their irrigation system or sprayer at that time. With the 7-day predictions, they can then choose when a more suitable time to perform that task will be, saving time and money if contracting third parties to do the work, argues Harvey.
“The overarching benefits to growers for having a solution like Sensing+ are convenience, time-saving, optimising farm resources such as water, chemicals, and labour, having a digital record at crop and field level to plan for each growing season, the ability to make evidence-backed decisions, and the 7-day microclimate predictions give them an unforeseen level of visibility over what is happening (and coming) on their farm,” says Harvey.
As well as collecting data from its sensor network, The Yield crowdsources some data from the farmers around growth metrics.
“We predict where a crop should be in terms of growth stage and will send farmers alerts to ask “is this where you are now? Yes or No.” Their response feeds back into our analytics to optimize the rest of the system,” says Harvey.
Eight weeks after deploying a sensor in a microclimate, The Yield is able to start offering farmers 7-day predictions based on conditions in their immediate environment. “You need enough of a data stream to reach an acceptable threshold for the system to self-learn and for us, at the moment, that’s eight weeks.”
Working with Bosch
While The Yield uses third party microclimate sensors, it designed the supporting hardware and mesh network needed to maintain connectivity across operations in partnership with Bosch, the German multinational engineering and electronics company. Bosch is also an investor in The Yield.
The Yield owns the IP on this enterprise-grade “farm area network,” which consists of nodes and a communications gateway, but contracted Bosch’s R&D team to help develop the design and manufacture them. The network is essential for the business because it means the startup can control the quality of the data its analytics platform ingests.
“One of our cornerstone customers is a prominent winemaker in Richmond, Tasmania, and they told us that every time a tourist bus goes past, their 4G footprint shrinks, which means they can suddenly lose a whole load of data. That is a big no-no for us; you can’t do predictions and build analytics from localized data if you don’t have a constant, high-quality stream. So, the area network is critical to avoid any gaps in the data stream,” says Harvey.
Later the company plans to build its own microclimate sensors too,
How Did The Yield Get Here?
“I often describe myself as the most unlikely entrepreneur; I’m a woman, I’m not a technologist, and I’m in my 50s.”
Harvey worked in international development for 15 years before founding The Yield, and while that might seem an unlikely beginning for a tech entrepreneur, her experience is very complimentary.
She founded a social enterprise called Better Work in 2002, which was a partnership between the United Nations and World Bank Group, which aimed to bring transparency to the supply chains of large clothing companies including Nike and The Gap. What ended up being a tech-enabled business, Better Work monitored conditions in these supply chains such as safety, child labor, and worker rights, with the earliest iterations of tablet computers and cloud-based data storage “although it wasn’t called the cloud back then,” says Harvey.
The business sold that information to the brands as a knowledge service.
“In international development, the holy grail is sustainable business models, and that’s how we approached using technology, as an enabler and facilitator, and not the end point,” she says. “Too often I’d see examples of people throwing money at problems, then the focus shifts and donors move on, and the initiative collapses. But if you come to technology in reverse, with the intention of solving a business problem first, and then building a platform to support it, it’s a much more sustainable and scalable approach.”
Moving back to Tasmania in the early 2000s, Harvey wanted to apply her thinking and experience to a new business problem and agriculture was a clear choice of a sector with many business problems to address.
Harvey also has great ambitions for The Yield to effect change to the food system and thereby contribute to reducing the impact of agriculture on the environment, and the company is already providing aggregated data to approved research organizations working on public good problems – with the express consent of their customers. But to make it work and scale, “we need to focus on execution; we won’t be able to realize larger ambitions if we don’t help farmers increase their yield through using our technology.”