I’ve attended a number of drone conferences over the past year, and while there are always especially noteworthy sessions, there’s one specific type which always generate a huge amount of interest and discussion. Those are the Venture Capital sessions, where a number of the experts who literally get paid to figure out what the future of the drone industry looks like come together to share some of their insights and experiences.
During a recent industry event, Jake Saper, Principal at Emergence Capital
, took the stage to take part in a discussion with other VC’s that focused on the commercial viability of drone technology, but also touched on what developments are going to be of critical importance as we shift into 2017 and beyond.
We’ve taken a look at what 2017
might have in store for the drone industry as a whole, but Jake was able to offer an even more refined look at how the future is shaping up. His company’s investment in DroneDeploy
is proof of their commitment to that future. In the interview below, he discussed some of his insights from the VC session, why software is going to be of critical importance going forward, what he thinks is going to dominate the headlines in 2017 and plenty more. Jeremiah Karpowicz: You recently participated in the "Finding Funding: The VC Community Speaks” session. What were some of your takeaways from the conversation you had with your fellow venture capitalist participants?
Jake Saper: One of the things I took away is that there is a still a diversity of perspectives amongst the funding community in terms of what strategy is ultimately going to "win" in the drone space. There was some alignment between us, but there were also some areas where there's a diversity of opinion.
On the alignment front, it was clear that everyone on stage agreed that we're in the first inning of the commercial drone ballgame. We're still extremely early, so any trends that are happening now are just starting, which means it's a little early to say, "this is the endgame". The flip side is that, as venture capitalists, we all get paid to make calls on what the medium and endgame are. That’s where opinions varied.
Bilal (Zuberi) at Lux seemed to voice the perspective that there will need to be a variety of hardware configurations to address a variety of enterprise use cases. Some of the other folks talked about how exciting investment opportunities were to be found on the services side of things. My perspective is that the most value over time will be created at the software layer. In what way?
My thesis is that as prosumer drone hardware continues to commoditize, it will become cheap enough that any field worker who wants to use a drone for their work will be able to own one. The hard thing will be to make the experience push-button simple.
I think we’ll get there though. My vision of the future entails a commercial user being able to buy a prosumer drone off the shelf, download an application, open that application on their phone, trace wherever they want to analyze and then watch the drone fly itself. The software will take care of collecting the sensor data, stitching it together into maps and analyses, and, ultimately, coach the user through whatever decision they are making with the data.
My bet, and I'm willing to admit I could be wrong, is that for the majority of commercial use cases, be it in agriculture, insurance, land surveying, photography, or anything else, a prosumer drone will be able to get whatever you need done faster, better, and cheaper. So we’re headed toward a future where drones are going to be commoditized in a very real way then, aren’t we?
Yes, but to be fair, there will be some commercial use cases where a prosumer drone won't cut it. Even with that in mind, my bet is that for the majority of commercial use cases, a prosumer drone will cut it, and as they get cheaper, the value will be created in the software layer.
The hardware parallel for me is the solar panel industry. I was in the solar industry before I was an investor, and the solar panel industry went through a really interesting process in the mid to late 2000s. The manufacturing of solar panels started in earnest around that time, and there was significant activity in Europe, and, to a lesser extent, the US. Panels were expensive and quality varied.
Then, Chinese manufacturers, helped by cheap debt from the Chinese government, built out really large manufacturing facilities for solar panels. What you saw was that during the late 2000s there was a huge amount of supply that came online as a result of these Chinese panels. That increase in supply, mixed with the financial crisis depressing demand, drove prices down in solar panels and actually made solar much more ubiquitous around the world.
While I see the parallels, I do think drones are further off from being fully commoditized. Today, there are meaningful feature and ease of use differences, but my bet is that over time there will be a variety of manufacturers making very high quality, high feature airframes. That will drive prices down just like in the solar market. I know that the panel identified the insurance sector as one where the biggest opportunities reside. What are some of the factors that make this sector so lucrative?
The very rough framework I use to think about where drones are most attractive uses a simple two-by-two matrix. The first dimension of the framework focuses on how big the space a drone is able to serve might be, and the second is how much value the tool can create in that space.
My perspective on the insurance industry is that first and foremost, it's really big. It ticks off that first dimension. Secondly, I do think there is a lot of value that can be created in the short and medium term.
Think about how much risk and potential injury goes into having people climb on top of roofs to do inspections and other types of work along those lines. Drones can prevent the need for humans to do all of that, which lowers liability and can perhaps even lower insurance costs themselves. There is a lot of value that can be created today just from an off-the-shelf prosumer drone. That issue and question around value is one I’ve seen come up over and over, whether it’s in terms of ROI or around how professionals can actually use these tools. How can the drone industry as a whole best address such things?
We have to think from the end-user perspective. What problem is the drone trying to solve? Who is it trying to solve it for?
If you're a farmer, what you want is not a drone. What you want is an answer. It can be an answer to questions around whether you have an irrigation problem, or whether or not there's a fertilization problem. If so, where? What is the topography of the land? These are questions they need to answer, and as an industry, we have to think about how we can help those end users answer their questions most effectively. And that goes back to the push button simplicity and importance of software, doesn’t it?
Absolutely. Just selling someone a drone, whether it’s an especially fancy one or just an off-the-shelf prosumer drone, is by no means enough. Even if that person knows how to pilot the drones themselves, unless you have an expert on hand to help sort through data, getting answers from the drone requires software.
As a specific example of that value, software can help you create an NDVI map, which is used in agriculture and is effectively like a green-scaling map. You can use software to autonomously fly a drone over a plot of land, collect data, and then stitch and map it all together in NDVI format to get a proxy for how much chlorophyll is in each area. You can use this tool to identify any brown areas which indicate a problem. You might not actually be able to see that with the naked eye or even raw images from a drone. That transition from “data” to “answer” is one that’s caused a lot of issues, hasn’t it?
Data for the sake of data is meaningless. There's no reason to collect data in a vacuum. You collect data to get an answer.
If I had one message to give to the drone industry, it would be to think less about the technology and more about what problems we’re solving for the end user. That's the key. Sometimes the best way to solve a problem is with an aerial drone, sometimes it's with an autonomous vehicle, and sometimes it’s neither of those things. Ultimately, it's all about finding the right tool to solve a specific problem.
One example of how we might need to adjust our thinking is around drone delivery. There's been a lot of talk about having UAVs deliver things like a pizza, and I do think we will see pizzas being delivered autonomously. But my bet is that in the majority of cases, those pizzas will be delivered via a land-based drone, aka an autonomous car, versus an aerial drone.
As an industry, we need to broaden the dialogue beyond classic conversation points like airframes and battery life. All of that is important, but I think we're now in a place where we need to start shifting our thinking to ask what problems can we help the end user solve. My perspective is that software will play an absolutely critical role in shifting that thinking. Will that shift result in a change to what users are actually doing with the info they collect from a drone, in terms of how they access and store it?
There are far too many siloed pieces of data in various industries. Thinking about how to get the data to integrate and build applications to do analysis is really important.
That's something DroneDeploy did when they opened up the app marketplace
. The marketplace allows you to fly using DroneDeploy and then plug in a series of apps to aid with analysis. Going back to the examples from the ag space, if you want to do some more specific analysis of your farm, users can plug their drone data into applications that are tailor-made to do that specific piece of analysis.
The reason that's exciting is because it allows a farmer to get the answer they need. Instead of just collecting the data and then storing the data in the cloud, you can plug it in to a specific app to figure out if something like an irrigation plan needs to be adjusted, or if they should change how they’re fertilizing an area. Apps that can answer specific questions after being seamlessly fed with drone data will create a whole new paradigm around the value these tools can create. That kind of ecosystem is clearly going to be a major development in 2017, but what else do you think is going to dominate the headlines in 2017 in the way that the announcement and rollout of Part 107 dominated 2016?
I think 2017 will be the year we begin to see the result of all the pent us demand that Part 107 has released.
At DroneDeploy, we had a bunch of enterprise customers that were kind of kicking the tires before 107, and now things have started to accelerate because enterprises can operate without fear of legal ramifications. My guess is that 2017 is really going to see the fruit of that policy change, and you're going to see much more significant enterprise players using drones.
Part 107 has provided clarity, which is just as important as the content itself. The idea that 107 gives project managers and risk managers within enterprises clarity around what's acceptable and what's not in terms of how they can adopt the technology is as important as the rules themselves. Do you think adoption will be hindered since Part 107 does not allow certain things like nighttime flying or BVLOS operation?
My thesis is that you start today where you can. That's line of sight, and I think there's a lot of low-hanging fruit within line of sight. Over time, as technology gets better and regulators get more comfortable, they'll relax the line of sight restrictions and more and more opportunities will open up.
There’s been a lot of talk around what Part 107 could and should allow, but now that’s going to transition into a discussion on what people are actually doing with drones out in the field. As we move to a future where we have less regulatory burden, the core challenge will be around how we prioritize solving problems for end-users.