Small drones are advancing quickly. They used to be little else but sophisticated remote-controlled toys and people love to complain when we call them ‘drones’ because that implies that they are at least semi-autonomous (and maybe able to bomb a terrorist (and a few innocent bystanders)). But what’s happening now is that the technology is catching up with the term. Drones are becoming increasingly autonomous and that means they will open up a range of new business opportunities for startups in the very near future.
As 3D Robotics CEO Chris Anderson noted during his keynote at the first annual InterDrone conference in Las Vegas last week, drones are now becoming “smartphones with propellers” instead of “airplanes without pilots.”
In an interview ahead of his keynote, Anderson told me that 3DR is investing heavily in research around autonomous flight, for example, and has hired a number of AI and machine learning experts to help it improve its technology.
Early drones were only as good as their pilots. Today, drones can perform some basic functions autonomously, but when it comes to avoiding obstacles, they are still pretty dumb. Even if most drones can now follow GPS points, after all, they’ll still happily fly into a wall if that’s what the operator who input those coordinates told them to do. And it’s worth noting that avoiding a wall or tree is a relatively easy task, but it’s much harder for a drone to see a power line that spans across a field, for example. The big technical obstacle for drones right now is this ability to avoid obstacles.
Once drones indeed become “smartphones with propellers”, they become part of the Internet of Things, but maybe more importantly, they will be able to talk to the other Internet-connected drones (and maybe planes) around them. That, combined with smart obstacle avoidance systems, is a necessity if we want drones to become more autonomous and open up new use cases for them.
Then, once a drone can work more or less autonomously, you can stop thinking about the complexities of operating it and start thinking in terms of the data you gather from its flights.
Anderson and many others I talked to at InterDrone believe that the current state of the drone industry is similar to that of the early days of the Web. That means we should now start thinking about what it means to combine this new drone technology with existing technologies — just like it was possible take virtually any existing industry during the early days of the Web and start thinking about how you could disrupt it by combining it with the power of the Internet (Web + restaurants = OpenTable, for example).
Some use cases are already fairly obvious. There’s precision agriculture, surveying, and — largely thanks to Amazon — delivery. I recently heard about a company that helps you stream imagery from drones into video conferences, too.
There are plenty of low-hanging fruits, but the most interesting ideas, of course, are the ones that aren’t immediately obvious.
With regulations still (slightly) in flux, there is still some uncertainty around drones and their use cases, of course. That’s not stopping venture capitalists, though, as Intel’s investment in Yuneec and the rise of a number of drone-specific funds clearly shows. While these funds are making their fair share of hardware investments, a lot of the development around drones is currently on the software side, too. All of these flights, after all, need to be planned and managed (and kept out of trouble with the FAA). That’s where the likes of Airware, Skyward and numerous other startups come in.
Indeed, what’s mostly happening right now is that a number of companies are putting the right infrastructure in place to allow for this drone + [x] scenario.
All of the data that drones will soon gather will also need to be analyzed, too. That’s pretty much a standard big data challenge and we’ll likely see a number of big data solutions for drone-specific use cases crop up in the next year or so. Farmers, after all, don’t have the time to analyze the imagery their drones will deliver to them every morning. All they need is a dashboard that tells them if their crops look fine or if they need to plan an intervention (and don’t think precision agriculture is the only use case for drones on farms — I also recently heard about a farmer who uses drones to scare off birds…).
Once we bring together truly autonomous drones that are aware of their environment, sensors and big data analysis tools (and sensible regulation), we’ll see if drones can live up to their potential.
Comment