July 3, 2016
How drones used to collect aerial data augment on-the-ground techniques and help a Chinese company better serve customers.
In May 2016, PwC launched a drone survey service in Poland. The Financial Times recently pointed out that countries such as Poland and Bulgaria, where 20 percent of the labor force is still in agriculture, are eager to make farming more cost efficient. This was an interesting concept, so I recently went to the AgTech Silicon Valley Conference at the Computer History Museum in Mountain View. I was hoping to get some more insight on the use of drones in agriculture – but more interesting to me was the concept of “drones as a data service.”
Drones as part of a larger aerial or aerial + ground data service offering make a lot of sense. Why would you own hardware, when what you really need is the right data at the right time and place? Many farmers at this conference said they were sick of buying information technology they had to operate and maintain themselves. It may not be financially prudent for farmers to buy or lease the hardware when the technology is still maturing and changing rapidly – not to mention having the time to figure out drone technology and complex data acquisition.
But drones aren’t the only way to collect data. In fact, aerial data collection on its own sometimes isn’t enough. The University of California at Merced, which had a booth at the conference, has developed a robot ground vehicle with a mounted camera for orchards. An orchard’s canopy, they point out, prevents airborne cameras from capturing what’s happening on the ground. If you want to study what’s going on in your orchards, you need a ground-based approach.
Mavrx, a VC-funded San Francisco startup, uses data from a variety of sources, including battery-powered drones, to help farmers monitor crop health. (Source)
I interviewed CTO Yuan Gao recently on the service his drone data services company Mavrx delivers. Yuan posts some very cogent and informed answers to questions about drone technology, power management and other topics on Quora, which is why I decided to contact him. During that interview, he dispelled a number of myths about using drones:
- Drones augment, but don’t replace existing survey technology. Manned aircraft have a lot more range, which is important when it comes to large-scale corporate farming, and imagery satellites have a lot of scale and overall utility. Companies like Blom, Mavrx, and PrecisionHawk use all these sources and then distill the information using data modeling, filtering and machine learning-style regression. These companies really are data collection and analysis service providers.
- Drone service providers don’t just focus on one kind of survey. Drone survey use cases are all over the map, from mining, to forestry, to disaster relief, to insurance claims adjustment, to plant, warehouse, storage yard, or agricultural use. It’s important to align with a company that handles the data you are interested in. For farming, it might be companies that primarily study the health of row crops such as corn, wheat, and soybeans, as well as grapes and other vine crops.
- Other drone use cases range from the mundane, to the challenging, to the highly specialized. Gao classifies use cases into five different areas: surveillance, survey, sport, cinematography and delivery. Surveillance is the most challenging because of privacy concerns. Delivery, he says, is a niche application—urgent medical delivery, for example. He doesn’t think ordinary package delivery using aerial drones makes sense in developed areas. “We have the roads, these nice, level surfaces that are well defined that we could have autonomous delivery trucks on.” Those roads offer the most energy efficient means for neighborhood package delivery, he says. To Gao, it makes more sense to use ground robots than aerial drones for the last mile if paved roads exist.
Do you have a story on the use of survey drones that you’d like to tell? Please post a comment if so. We’ll be publishing more on drones soon, so stay tuned.