The International Potato Center (CIP) is working to improve data collection while driving down costs by building tools that gather vital information without ever touching a plant. Driven to improve the quality of information, an innovative team at CIP has pushed themselves to build their own equipment designed to help them streamline the data captured.
By studying the interaction between plants and incident light, scientists can investigate everything from how the changing climate is affecting the plants, to how they’re reacting to pests and disease. This type of work, known as remote sensing, is the art of measuring without touching, and allows scientists to gather information on a massive scale. The technology can alert researchers to crop stresses, helping farmers make more informed decisions and plan for markets, or cushion themselves against the impact of drought or disease.
Hildo Loayza, research associate at the International Potato Center, shares with us five tools scientists are using to build equipment that deepens their understanding of plant behaviour.
3D technology has allowed us to dream and bring forth new ideas. We can now build what we envision. In the past, we were forced to adapt to the tools available. Commercial tools are designed to be used by the widest audience possible. The software is created to gather data sets based on the parameters determined by the manufacturer, so it has certain limitations. At CIP we are focused on studying very specific plants, roots, and tuber crops. We know how high potato plants grow, we know how deep sweet potato roots reach. With 3D printing, we can design tools adapted to our needs. We’ve used it to build parts for our drones and our cameras and sensors. This not only allows us to get more precise data, it also reduces costs. Our 3D printer itself has even become self-replicating. As parts wear out, we simply manufacture new ones.
CIP has a long history of collecting remote data. Our group leader, Dr. Roberto Quiroz, was known for using hot air balloons and electric and combustion airplanes to capture crop information. These were unwieldy and hard to navigate. In 2012 we began using multi-rotor drones which are much more precise and easier to control. The drone itself doesn’t gather data. It is simply a mechanism that helps transport the tools we’ve developed for data collection.
The easy navigability of a drone (as seen above) means we are able to get much closer, more detailed information of large land areas, than we could with a satellite image, at a much more affordable price. Such information helps scientists better understand how crops are evolving in a particular area over a prolonged period of time. We adapt our drones to carry the camera equipment we’ve designed to maximize our understanding of roots and tuber crops.
While there are pre-made cameras on the market, they are expensive and built to capture a lot of information that we don’t need, often at the expense of image quality.
At CIP, with the help of lenses and filters, computer parts and our 3D printer, we’ve essentially developed a Frankenstein type camera called the IMAGRI, which stands for Integrated Multispectral Agricultural camera system. The light reflection wavelength of each variety of plant differs. We build the IMAGRI to hone in on these differences and capture much more targeted data. A commercial camera, without the adaptations needed for fitting to a drone, runs to roughly $3,800, while a fully adapted camera built at CIP registers at less than a third of that cost, at roughly $1,100.
Plants offer us a wealth of information that cannot be seen by the human eye. Using near infrared (NIR) imaging we can observe how a plant is reacting to certain stresses with much more accuracy. It’s like when a person has a fever. I might not be able to tell by looking at you that you have a fever, but you’ll still feel it. NIR helps us pick up that something has changed from what’s normal for a plant, just as a thermometer might tell us a person has a temperature. We set a baseline so that we know when the plant is doing well and when a plant is feeling stressed. We then use NIR imaging and a combination of spectral bands to build vegetation indexes and monitor for changes in a plant. The information we gather is useful to farmers. It can let them know when to irrigate, how much to plant, or even when a farmer should pass on planting a particular variety because it is not suitable for that area.
Open Source Software
Making our cameras function, and stitching together the images collected, requires software. Our team relies on open source tools that we program to run our cameras and interpret information. Taking a series of photographs over a large area that is later stitched together gives us a more comprehensive picture of what is happening on a larger scale than we could get with an individual image. One of the limitations we had with pre-built cameras was that the programming sacrificed image quality in order to capture data that didn’t particularly suit our needs. Since we have very defined parameters for our crops we can program for them and get a better image quality as a result.
In the end, the engine of innovation is the human mind. A 3D printer, a drone, or a camera offer nothing to deepen our understanding of plant behaviour without the person operating them. It is the person who innovates. The scientists who see the connections between the data being gathered and how this can help farmers get the most from their crops. The CIP team is made up of a mechanic, a physicist, and one or two electrical engineers on any given project. We build on one another’s strengths and rely on our collective decades of experience on roots and tuber behaviour to develop the tools that can help get smallholder farmers the information they need to make the most of these crops.
Hildo Loayza is a research associate with 11 years experience working at the International Potato Center. He is currently working towards his Ph.D. from Pierre and Marie Curie University and is conducting doctoral research at CIP in the use of chlorophyl flouresence to detect water stress in potato and sweetpotato. He is based out of Lima, Peru.