On the use of drones… and their limitations

We’re starting to see many entrepreneurs and major companies rushing into the world of (non-military) drones, as these little robots promise to solve any problem, any where, any time, automatically, maybe after a short research and development phase. But how short is this “short research and development phase”? Or, more importantly, what are the major limitations of these platforms?

Some might argue that the major bottleneck (specially those drones enabled for vertical takeoff and landing – VTOL) is the on-board power, which restricts the flight endurance, the payload, as well as the communications’ range and bandwidth. There is a myriad of emerging new battery technologies that promise to bring about a major revolution in this area. Nonetheless, these promising batteries technologies are still in the research phase and it may take a decade (I’m an optimist) to have one of these to be mass produced, in order for prices to drop to those of the current Lithium-polymer and similar technologies.

Another set of important issues are related to people safety, privacy preservation, insurance requirements, and general law compliance of the drone and its operator. There have been already plenty of accidents involving drone failures or drone operation by unqualified personnel (either in terms of technical knowledge, ethical standards, or safety measures), which resulted in injured people and some nearly collision with commercial flights. As a consequence, and to prevent more bitter consequences, legislators in EU, USA, and other countries are moving to ground any drone that does not have an explicit authorization from the airspace regulator, while working on more comprehensive regulations to integrate them in the usable airspace.

Finally, the most difficult challenge (in my opinion) is to enable any drone to have a thorough perception of its surrounding environment. To the best of my knowledge, there is no commercially available (non-military) drone capable of localizing itself and have a representation of the surrounding environment. This means that no drone is capable of navigating through a city, to deliver pizzas or tacos or books, avoiding buildings, cars, people, or other drones, particularly if these are moving obstacles. This difficult problem, along with other technical problems that are currently under intensive research, is not acknowledged by most of those going into this industry, and may frustrate some, otherwise, great ideas for a business plan.

To sum up, before thinking of starting a company involving drones (or, really, any other technology) one should have a clear understanding of what they can do NOW, and what are the perspectives of having those enhanced capabilities in the near future.

My Arducopter Hexa – Entering the T3 Cube Challenge

As a New Year’s resolution, I decided to invest in a new hobby: Drones.
In particular, the Arducopter Hexa, by 3DRobotics!

Well… I also do research on the use of drones for automatic inspection… Nonetheless, in between observability proofs, filter/controller design, Matlab simulations, and writing papers, I really don’t get that much “hands-on happiness” as I’d like to have! When we have a very expensive drone equipped with thousands of Euros in payload (lasers, cameras, embedded computers, and so on), only professional pilots (unlike me) should handle them!

I bought the Arducopter Hexa B, with uBlox LEA6 GPS, telemetry, and sonar. During the weekends, I started with the basic Attitude mode, with some hover and basic flight maneuvers. As I gained more confidence in my abilities as a pilot and on the behavior of the vehicle, I started exploring the Altitude Hold and GPS modes, which despite the very noisy sonar data (checked in the logs) worked very smoothly. I eventually started exploring the return-to-launch (RTL) mode, and finally, my goal, the Auto mode where we define our mission in the Mission Planner and the Arducopter executes it when we switch to Auto mode during flight.

At some point DIYDrones.com put in place the T3 Cube Challenge:
“The mission is simple, get airborne climb to 20m.
Fly a cube with 50m sides, pausing for one minute at each corner, so your flight time cannot be anything less than 8 minutes….. Bonus points if you can stay longer at each corner….
The neatest cube KML wins.”
This was the extra push I was needing to get deep into mission planning!

After four missions, three of which were cube trials, I managed to draw something similar to a cube in the skies of Portugal. The drone remained aloft for 13 minutes, spending around 25 s in each corner, but going through some edges twice, to ensure all edges of the cube were visible in the final trajectory.
The image below shows the result of the mission log with which I entered the T3 Cube challenge!

Not bad from my perspective, but there were more accurate cubes and far more longer flights, which got the top three places and prizes. There were eight people from around the world participating in this challenge and, just for entering this challenge, the kind organizers offered me a t-shirt! 😀

The KMZ file can be found here.