Stand in a cornfield in Iowa, a shrimp farm in Vietnam or a palm oil plantation in Malaysia and look up. 300 miles above you, a battery of satellites the size of shoeboxes is looking down, relaying data that can be used to determine everything from crop biomass to water transparency.
Launched into orbit as secondary payloads by rocket makers, Planet’s satellites are loaded with high-resolution cameras that photograph Earth’s surface daily, enabling near-real-time monitoring of farmland around the world.
Its ‘SuperDove’ satellites have a 3.7-meter resolution on a near-daily cadence, which sets them apart from public satellites such as the European Space Agency’s Sentinel and NASA’s Landsat satellites, which provide free imagery but at lower resolution (10–30m) with less frequent coverage (every few days).
The latest generation of tasking satellites (‘Pelican-2’) provides data across six spectral bands, with a resolution of 50cm, and further down the road 30cm, providing far more granular data.
The next generation product—the ‘Owl’—will have a one-meter resolution, faster analytics, and advanced onboard AI via Nvidia GPUs, with a tech demo slated for launch late next year.
While its satellites can’t see through clouds, with its Crop Biomass product, Planet can fuse its PlanetScope optical data with radar data from other satellites (which send microwave signals that penetrate clouds) to get continual coverage. Similarly, given that they circle the earth daily, they have more shots at target, notes Lead Product Manager, Agriculture, Ariel Zajdband.
“We have a lot of pictures of clouds! But the higher the frequency of collection, the higher the likelihood of collecting a cloud-free image, let’s say within a week, right?”
From data to actionable insights
The challenge for Planet and every company collecting vast amounts of data in agriculture, he says, is turning it into actionable insights.
In the case of farmers, says Zajdband, most are probably unaware that they are utilizing Planet data as they access it via the digital ag platforms of seed and crop protection giants such as BASF and Syngenta.
“Most of the farmers who use PlanetScope data don’t know about us because they consume our data through our partners’ digital platforms. But we do sell our data directly to some large farming operations, for example large soybean growers in Brazil or sugarcane growers in Latin America, who have their own remote sensing teams and their own systems that they build. In these cases, they just connect our data to their own pipelines through APIs.
“But we’ve also been working on making it easier to use with what we call ‘analysis ready data’ or ARD products that a data scientist can use. You don’t need to be an expert in remote sensing to work with satellite data.”
How AI can democratize access to satellite data
In the past, “the old paradigm of remote sensing was that you had to download a lot of huge files and build your own server or pipeline,” says Zajdband. “We built our own cloud-based platform [Planet Insights Platform] that allows our customers to process all this data in the cloud and extract just the insights that they need.
“We’re lowering the barrier to entry so we can have more users who don’t need to be experts in remote sensing.”
Planet Insights Platform, which has several tools to help customers develop apps that process earth observation data, also has APIs and SDKs (Software Development Kits) enabling them to integrate imagery into their front end web applications, he says.
While Planet has renewed its focus on AI, many of its customers are using AI to help them gain insights from satellite data “much faster and directly,” adds Zajdband.
In future, he says, AI tools may mean that users without coding skills or even data science skills will be able to gain insights from Analysis ready data products such as Planet Fusion or the Planetary Variables. “This will lower the bar even more and enable more users to [gain insights from] our data.”
Precision agriculture
But what kind of data is most relevant to the ag sector, and how is it being used?
“People accessing our data via digital farming platforms are using it mainly for two things,” says Zajdband. One is scouting fields for diseases, pests, weeds, or water stress. The other is generating variable rate applications using satellite data for things from seeds to fertilizer to fungicides.
“By understanding how crops grow differently across a field, you can apply different rates of different inputs throughout the field.”
Abelio, a precision ag company in France, for example, has developed a crop monitoring solution utilizing PlanetScope imagery to help farmers tailor nitrogen application to the specific needs of each zone within a field, enabling farmers to reduce nitrogen use by 5-10% on average.
Planet’s Land Surface Temperature product meanwhile, measures surface temperature using a blend of cloud-penetrating microwave and infrared/optical data enabling customers to track crop stress, drought, and heat patterns and model growing conditions more precisely than air-temperature data alone.
This could be used for yield prediction in combination with other data such as Soil Water Content; for helping growers determine the best depths for seed planting in different parts of their land; or for predicting plant and pest development rates, says Zajdband.
“If you can model the risk of developing a given disease or pest, it can help you to decide to apply it before [the problem develops]. And then once you decide to apply it [a crop input], you can use our data to measure how much crop, how much biomass, or forecast yield, and say, OK, in this part of the field, is it worth applying fungicide or not? At what rate?”
Soil water content, pasture management, aquaculture
Others use PlanetScope data to look at soil water content, says Zajdband. “In this case, we are using data from other [providers’] satellites to provide data for governments and private companies supplying parametric insurance to farmers.” In the case of drought parametric insurance, for example, payments are triggered if the amount of water in the soil falls below a given threshold.
“We also work with customers who are using our soil water content data to understand soil trafficability; are the soil conditions good enough so I can enter with a sprayer and apply product without causing compaction?”
Another area of steady growth, meanwhile, is around the application of virtual fencing for pasture management, he says.
“Our data can be used to measure forage availability in a pasture [helping virtual fencing companies know where best to direct cattle]. Historically, we’ve been working mostly with companies in New Zealand and Australia on this, but now we have companies in many different regions using our data to measure in near real time so they can optimize forage.”
Another emerging use case is in aquaculture, he says. This is both for monitoring water transparency and turbidity in ponds or coastal water for early detection of water-quality changes or for mapping salmon farm locations, cage counts, and surface area, helping monitor aquaculture expansion, compliance, and environmental impact over time.
Validating regenerative ag, managing farm payments
For some institutional customers, says Zajdband, PlanetScope data is used for compliance purposes, whereby governments issue payments to farmers based on the extent to which they adhere to environmental or other rules.
“Each payment agency has its own set of rules, which might be around tillage or a crop type, or mowing events. In this case, we are not only collecting data with our satellites but providing the tools and the digital platform for our customers to determine how to distribute subsidies under the Common Agricultural Policy in the EU, for example.”
Nitrogen status monitoring, a ‘transformational’ new use case for hyperspectral imaging?
Moving forward, while the first generation of Planet’s satellites conduct multispectral imaging, its new Tanager satellite delivers hyperspectral imaging, capturing 420+ spectral bands with 30-meter spatial resolution, allowing for finer discrimination of materials and surface properties, says Zajdband.
“Right now we have one Tanager in orbit, and we are building three more. Having narrower bands allows us to measure things we cannot measure with multispectral, such as methane emissions.”
“We are also very interested in understanding nitrogen levels in the crops, a use case we’re starting to develop that we think could be really transformational for agriculture.”
“The nice thing about this particular use case is that you don’t need global coverage of every field every day, just a few times in the season when you can use that data to make a decision as a user.”
Smallholder farms
Stepping back, he says, “Often when people talk about use cases in agriculture, they often relate to big agriculture, like sugarcane and soybeans.
“But our data can also be used for smallholder farmers that work with a lot of the big crop protection companies we supply data to, so we are very interested in increasing our presence there.”
>>Learn more about Planet Insights Platform and get a 30-day free trial.
The post From orbit to insight: How Planet powers the next wave of precision ag appeared first on AgFunderNews.
- Source: https://agfundernews.com/from-orbit-to-insight-how-planet-powers-the-next-wave-of-precision-ag
