Dave Kivioja has always been fascinated by aeronautics, having spent the better part of his career flying airplanes for the U.S. Air Force and Northrop Grumman. Over the years he has also developed a growing interest in drone technologies and their applications.
After many conversations with a good friend, he realized a unique opportunity where he could apply his flying expertise to help farmers tackle a major problem – weed management.
“One of my buddies owns farms where he grows tons of corn, soybeans, wheat and barley. We had many conversations over the years where he’d talk about what a problem weeds were, especially Johnson grass,” said Kivioja. “I’m a problem solver by nature, and one day I got an idea – what if I could use drone technology and create NDVI maps that would help him identify where the Johnson grass was growing?”
Kivioja quickly found out that even with a drone, stitching together images and manually mapping out weed locations was a labor-intensive endeavor. He knew there had to be a better way, and that artificial intelligence (AI) could provide the solution. Further, he knew that identifying weeds was just one part of the problem.
“I not only wanted to utilize AI to make it easier to identify weeds in a field, but I wanted to develop a method to be able to spot spray them with herbicide,” said Kivioja. “Most drones out there do broadcast spraying, which is not only more expensive than spot-spraying but also results in more runoff into the ecosystem.”
He added, “I needed more funding to be able to do this, though – that’s about the time where I sought additional support from TEDCO.”
TEDCO, Maryland’s economic engine for technology companies, is known for the innovative programs and funding opportunities that promote the growth and diversification of Maryland’s entrepreneurial innovation ecosystem. And, in a stroke of luck and perfect timing, TEDCO launched an initiative that could support Kivioja’s problem-solving journey—the Agriculture and Rural Rebuild (ARR) Challenge.
TEDCO’s ARR Challenge was an initiative launched in early 2021 with the purpose of supporting rural and agricultural businesses in Maryland that had been adversely affected by the ongoing pandemic and economic downturn. These grants, which provided up to $200,000 in funding, were designed to encourage collaborations between applicants, research institutions, and industries. By providing this foundation, TEDCO’s vision was to facilitate the development of long-lasting technological improvements in the Agtech field and other rural industries. Eagle Eye Imaging was one of seven award recipients.
With fresh funding in tow, Kivioja quickly got to work on the AI portion of the project. “Creating georectified orthomosaics is common in the drone industry. However, using AI to quickly identify weeds was, at the time, still in its infancy,” noted Kivioja. “I started training the AI by teaching it what certain weeds look like against various backgrounds. You also need to teach it what areas of the field don’t have weeds so that it can distinguish. It’s a lot of work in the beginning, but after a while it can function pretty well on its own.”
Of course, getting the AI to work correctly was just one piece of the puzzle; Kivioja still had to get the drone to precisely spot spray the detected weeds. “The drone can only spray about a 10-foot circle diameter, but GPS can be off by about 15 feet,” he said. “That was a major kink that we needed to work out – it took about two years to fully develop the technology for the spot spraying.”
The hard work has paid off, though. “We found that with our spot spraying drones you’re only using about 10-20% of the amount of herbicide you’d typically use for broadcast spraying,” said Kivioja. “That means we not only save these farmers money, but we also help the environment by not putting as much product into the soil.”
“Broadcast herbicide spraying, while convenient, can pose a major threat to nearby wildlife and water systems,” said Dr. Arti Santhanam, executive director of the Maryland Innovation Initiative and administrator for the ARR Challenge. “Dave was able to apply his expertise in aviation to come up with a better, more sustainable solution to tackle this problem. His work exemplifies why we started the ARR Challenge – to drive innovation in agriculture for a better tomorrow.”
Since the initial grant project, Kivioja has continued pursuing opportunities to test the technology. Last summer, Kivioja did a field trial with two different drones at Texas A&M alongside Dr. Dan Martin, a USDA aerial application expert, where they were able to show that Kivioja’s technology was just as effective at detecting and spot spraying weeds compared to more expensive alternatives. Kivioja has also continued visiting local farms to further validate the platform and build confidence with farmers utilizing the product.
“Broadcast spraying is what farmers have been doing all their lives – I can empathize [and understand that] if you’re asking them to use a new technology that it better work well. I want to go to more farms and show them what we’re able to do so they feel confident in making the switch.”
As part of his plan, Kivioja is also getting out into the community to teach others what he has learned over the years. “Drones have a lot of potential for agriculture, but they also have a lot of FAA regulations behind them. If you’re not already trained in flying, it can make your head spin,” he said. “I’ve taught multiple classes for the Maryland AG Extension Office in Howard County where I focus on FAA rules and spot spray technology.”
“Whether I teach a class or go on the farm, people get really excited about this technology and its potential. I want to continue to get the word out because I know firsthand how much of an impact this can have in agriculture.”
Source: The Daily Record