Agtech startups help farmers harvest the power of data in the cloud

Think about farming and you will most likely to think of images of livestock grazing or crops growing in a field. But in modern farming, every one of the hundreds of actions that take place each day creates data. And when that data is harvested and analysed, it provides a wealth of information that is benefitting New Zealand farmers in countless ways.

With Fieldays 2020, New Zealand’s largest agribusiness exhibition, in full swing this week and taking place virtually for the first time, it is a good time to reflect on the benefits that digital technology is delivering to New Zealand’s agricultural sector.

Across New Zealand we are seeing examples of startup customers using AWS to help meet the challenges of farm and environmental management and maximise their return on investment. New Zealand already boasts one of the top 10 global ecosystems for agtech, as measured by the Startup Genome: Global Startup Ecosystem Report 2019, with 20 per cent of very early stage startup investment going to the agtech and new food sector since 2013.

We are pleased to support farmers, growers, ecologists, and other customers who serve the agricultural industry by creating new cloud-based offerings that can help them make better decisions from data, and much of this work is brought to life by innovative agtech startups.

Data-driven dairy cows

Craig Piggott has a close connection to New Zealand’s dairy farming industry, having grown up on a farm in the Waikato region in the North Island. After completing a degree in mechanical engineering and gaining his first job in a startup, Rocket Lab, Piggott knew it was inevitable that his mind would turn to how he could use his skills to help the agricultural community he knew so well.

That led him to think about how sensor technology could be used to better monitor the movement of cows across a pasture, and how it could help to guide their movement.

The result is Halter, a startup that created a sensor device that can be worn on a collar around a cow’s neck. Not only can the collar record data such as the cow’s position and movement, but it can also generate vibrations which can be used to encourage the cow to move to specific locations, such as to different areas of a paddock for feeding, or away from environmentally sensitive areas such as rivers.

“Labour is the number one challenge on a dairy farm,” Piggott said. “So the ability to manage a heard remotely from a smartphone is something that can deliver them significant productivity gains.”

“For example, rather than having to direct cows to the milking shed first thing in the morning, a farmer can use technology to guide the cows there, and complete other jobs instead,” Piggott said.

The data these devices generate also provides enormous benefit regarding the health and fertility of cows. Monitoring the cow’s activity through the collar provides clear information regarding her fertility cycle, and other signals can show when she is about to calve, or if she is having calving trouble.

“There is a small window every 21 days when they are fertile, and the more accurately you can predict that window then the higher chance they have of breeding,” Piggott said.

Halter can also help farmers to identify if cows are injured to ensure they can respond to their needs. “Cows are stoic, so they hide their injuries. We can detect changes in behaviour and the onset of physical problems weeks before a farmer can tell you visually. That enables you to go straight to the cow to help them a lot quicker than you otherwise would.”

Halter uses a range of AWS compute, data storage, and IoT analytics services for crunching data from the collars and outputting it as alerts and advice for farmers.

Halter is currently engaged in early on-farm deployments across Waikato and beyond.

“The opportunities to improve how the industry works are really cool,” Piggott said. “Every farm is so different, so we are still learning a lot and applying that to build on our existing solution.”

Protecting the environment

The importance of data for monitoring the health of the natural environment was also the inspiration for RiverWatch, a startup that is using waterborne drones to monitor the quality of New Zealand’s inland waterways.

RiverWatch Founder, James Muir came up with the idea after seeing the health of a river near his father’s farm deteriorate due to unsound farming practices further upstream.

“We realised we needed to get more data to show government stakeholders, and encourage landowners, farmers, and community groups to take their own action to improve waterways,” Muir said.

“So we came up with the idea of a real-time monitor that can you can have anchored in the river and send data consistently back to you so you always know what was happening in the water.”

RiverWatch debuted its prototype floating river drone at Fieldays in 2019 and was swamped with interest. It also picked up an AWS Innovation in Data award at the event, which Muir said gave the team at RiverWatch a direct connection into the AWS team.

“We started working on integrating everything we had been building and getting set for a future on AWS, using emerging technology like machine learning to make RiverWatch smarter and able to deliver high quality intelligence on our water,” Muir said.

The RiverWatch drones incorporate low-cost sensors that monitor a variety of factors relating to river health, such as oxygen levels, turbidity, and water temperature. Because the devices are in constant contact with the water, they can quickly detect changes, while also building up a thorough picture of river health over time.

Thanks to the analytics capabilities provided by AWS, this data can be used to provide advice on river management and can predict future conditions and advise remedial actions ahead of time.

“For example, we can look at what is happening in weather patterns and rainfall and make predictions for farmers based on the conditions they are likely to see,” Muir said.

Delivering value for farmers

Data plays a key role in helping farmers manage their stock and land, and it can also help them ensure they are being fairly compensated. This is the goal of digital auction management startup, bidr, which is using cloud-based video streaming and collaboration tool that runs on AWS to create an online virtual saleyard for trading livestock.

bidr’s general manager Tania Smith describes the service as providing a ‘saleyard dynamic’ and the benefit of an auction, but with the advantage of a farm-to-farm transaction.

“Animals don’t leave the farm until they are sold which supports animal welfare and benefit the environment by reducing trucking.” Smith said.

bidr is building up a strong dataset from more than 200 auctions that it has hosted to date.

“We’ve sold over 50,000 head of stock, and are able to generate statistics for our vendors and agents,” Smith said. “We use this data to promote and market to vendors, especially when we can tell a positive story about price premiums.”

Over the course of May and June 2020, bidr hosted 48 two-year-old bull sales, of which 31 were livestreamed. bidr provided clients with detailed reports about buyers, watchers, and bidding patterns. The most successful sale saw 30 per cent of bids and bulls purchased by online buyers. As all lodged bids are recorded in bidr’s database, the company can gain a picture of the type of animals that are of interest to different buyers across the country.

“This information is highly valued by listing agents and genetics vendors in particular,” Smith said. “Over time, we will look to compare the prices achieved in bidr auctions versus other sales channels, such as saleyards and at the meat works.”

“It’s been hard work delivering a new innovation to a 150-year-old industry. But in other markets around the world, online trading of livestock is becoming more prevalent, and when you enable the trading through an auction you are maximising the outcome for vendors.”