In The Air: Doing It With Agricultural Drones
From the service provider standpoint, getting into the aerial imagery as a service game was once an either/or proposition: Either you partner locally with a fleet of aerial application planes for weekly flyovers with a sensor payload, or you purchase a subscription as a provider of one of the many satellite-based imagery programs out there.
And it should be noted that there is — and remains — nothing inherently wrong or less effective with plane- or satellite-based aerial imagery services. In fact, for some situations where economic scale (i.e, 1,000-plu acre fields) or weather issues (high winds) make drone flights a tough day-to-day prospect, those methods of sensing can oftentimes prove preferable.
Yet with the ongoing advent of dramatically cheaper, easier-to-use unmanned aerial vehicle (UAV) technology capable of conveying agronomically powerful, real-time sensing technologies from the edge of virtually any field (or the bed of a pickup), at any time, almost anywhere (drones don’t have to be “tasked,” or scheduled days or weeks ahead of time, like satellites or manned craft), it can’t be denied that there is opportunity for service providers to integrate drone-based imaging to demonstrate a higher level of service with grower-customers.
Back in early-January, Sentera, a Minnesota-based ag tech startup that outfits UAV platforms with professional-grade sensing technologies, announced its newest advanced sensing offering to the precision farming market: livestreamed Normalized Difference Vegetation Index (NDVI) data visualization via the company’s Double 4K Sensor, otherwise known as LiveNDVI.
The ruggedized Double 4K sensor integrates with either Sentera’s Omni quadcopter or the Phoenix 2 fixed-wing drone (for larger acreage flights) and in-flight streams real-time NDVI data, which is used to measure and quantify plant biomass at various stages of growth. Agronomists or crop consultants on the ground armed with an internet-connected mobile device can then immediately use the visualized data for directed scouting services.
For context, most competing NDVI imagery systems typically require a 24-hour post-flight waiting period for the imagery to be geo-stitched in post-processing, after which point the image is pushed to the cloud and then delivered to the appropriate platform for visualization. Sentera has streamlined that somewhat cumbersome process by pushing most of the post-flight computing power and image-processing algorithms onto the sensor itself, rather than waiting to do it on a desktop back at the office). It’s an emerging trend known throughout the IoT world as “edge computing,” or as Greg Emrick, Executive Vice President, Business Development, describes it, “moving those (data) analytics upstream.”
Emrick sees the new capability as an “efficiency and productivity improver” for the precision agriculture professional already tasked with serving and stewarding five- or oftentimes six-figure acreages on an annual basis.
“We’re the only NDVI sensor producer that has the ability to process data on the sensor and produce a result in real time,” he explains. “That allows those agronomists or consultants to not have to walk the whole field; now they only go to the relevant spots that they need to look at, and it just reduces that time that it takes them to help someone make a decision.”
Sentera’s UAVs are currently the only drone-based aerial sensing platform that is fully API-integrated with John Deere and the MyJohnDeere Operations Center, and the NDVI data — usually spit out in the form of a heat map — can flow into myriad farm management information systems. Emrick lists SST Software, Ag Leader SMS, ADAPT-N, John Deere’s APEX, EFC Systems, Farm Works, and Pioneer Encirca when asked for specific software platforms that will process the real-time NDVI imagery.
“That professional agronomist might have 25,000 acres annually under his advisement. Now with a UAV he can maybe double that amount of acreage, while still providing the same level of service throughout the season,” he points out. “The other thing I think crop consultants will find is, there’s a fascination with drone technology that will at the very least earn them an opportunity at the table with that grower.”
Multispectral with MicaSense
Seattle, WA-based MicaSense also markets a drone-mounted sensor for precision agriculture professionals with its RedEdge five-band multispectral sensor. RedEdge consists of a downward-facing, five-spectral-band capturing sensor that mounts on virtually any drone, as well as a smaller, top-mounted and upward-facing “irradiance sensor” that allows the post-processing software — in this case MicaSense’s Atlas image processing software — to compensate for sunlight and reflections that can cause false readings in the imagery.
Gabriel Torres, Founder and CEO, says the key difference in what a grower-facing agronomist or consultant can do with RedEdge imagery compared to your standard NDVI imagery is in helping the service provider see and quantify what’s going on inside the plant, and then using that data, combined with tissue and soil sample data from the ground-truthed areas, to provide management recommendations that go beyond what one can see with the naked eye alone.
“There are some limitations with NDVI — it’s useful for measuring leaf counts and quantifying the amount of green leaves in the canopy, which is good for plant health status — but the combination of the red edge band plus other multispectral bands allows us to generate outputs that are indicative of chlorophyll in the leaves” ” Torres explains.
That current snapshot in time of plant chlorophyll content can be useful in many ways, including early identification of chlorosis, which is a deficiency of chlorophyll in the plant leaves that can signal a host of potential agronomic problems, and the service provider armed with an Atlas-generated chlorophyll map can base several management decisions off the data, according to Torres.
“We can use that data to guide nutrient applications — chlorophyll content is much more closely correlated with nitrogen uptake than NDVI alone, and so for cropping systems using variable-rate nitrogen it is a more accurate way of determining which parts of the field are likely to need more nitrogen and which areas won’t.”
Torres is quick to caution that RedEdge captured imagery data is not a “magic wand” and that the role the agronomic service provider plays in the data interpretation process will continue to maintain a high level of value, even as the sensing technology increases in sophistication and automation going forward.
“I think it’s important for service providers to understand the technology in general goes beyond drones and beyond sensors — that the value is in the combination of the ability to capture multispectral data with a good quality sensor and with the proper workflow that correctly optimizes and processes the data. Then that is combined with the expertise of an agronomist or crop consultant who can generate a prescription or action based off the data that we provide. That’s where the true value proposition comes into play.”