A Rational Approach to Regulating Private Drones

Larry Downes of Harvard Business Review published last week an article about seeking sanity in how to approach regulation of private UAVs.  So far, it’s one of the best I’ve read–pointing out that regulation is needed, but too-hasty and too-restrictive will likely require backtracking.  A couple of nuggets:

  • Some of the things that UAVs can do that legislators are afraid of, are already illegal to do with a UAV or without.
  • “Creepy new technologies” have been suspected of breaching personal privacy for a long time now.  In 1890, that technology was short-exposure photography.  Is anyone unsure whether we figured out how to deal with that?

A Rational Approach to Regulating Private Drones | Harvard Business Review

UAV + College Football = Ba-Bam!

I don’t know if Tennessee is ever going to turn around their football woes, but Butch Jones already chalked a W in my book. Combining two of my favorite things nearly brings a tear to my eye.

Let’s think about this: football coaches could spend a few hundred dollars on a quadrotor with an HD camera and get a shot of his players from nearly any angle in any position on the practice field, changing positions on demand. Or, he could spend thousands — maybe even tens of thousands — on an rig or camera platform that lumbers around and can’t get every shot you could want. Hmm.

Tennessee uses a quadrotor to give coaches better perspective | TechCrunch

StopRotor Hybrid Rotorwing?

I saw a little nugget on Gizmag the other day about a new design an Aussie is working on for the DARPA X Plane program:

StopRotor Technology’s Hybrid RotorWing can take off vertically and hover with the main rotor spinning like a helicopter’s, or take off and fly like a conventional fixed wing aircraft. To transition between rotary and fixed wing flight modes, the aircraft is plunged into a fall involving high angle of attack flight that aligns the airflow with the rotation axis of the rotor system. This symmetrical airflow over the main rotor provides a stable environment that allows the blades to be started or stopped.

Here’s my take on the good, the bad, and the ugly for this.

The Good:I’m a big fan of innovation and novel flight concepts.  This designer definitely has some creative ideas about how to do fixed- to rotary-wing transition.  I don’t think this thing is going to work, but it’s worth a shot to investigate.

The Bad: The reason I don’t think it will work is because a) the concept will only work on a UAV (see The Ugly below), and b) there are cheaper, much more practical ways to have effective VTOL capability with a fixed wing UAV.

The Ugly: This is (I think) a novel idea to rely mostly on aerodynamic braking and acceleration to start and stop the rotor, but the simulations shown in the videos make me shudder.  When that rotor really slows down, but hasn’t yet been locked in place, the craft will be VERY hard to control and probably highly unstable–who knows if it can gain enough initial altitude to complete the transition without crashing.  Second, those rotor blades/wings are so skinny, it’s difficult to imagine them being stiff enough as the craft is scaled up, not to mention the stall characteristics.  Finally, the pitching and rapid descent profile for transitioning the rotor looks like it would be a vomit-fest for any person riding on board such a vehicle.

Stoprotor Technology Pty Ltd HybridRotorwing

Bell 407GX First Civil Helicopter To Land at Hongqiao

Image courtesy Gerry Metzler via Wikipedia. Licensed under Creative Commons license.

In an awesome sign of change for the international aerospace industry, and the rotorcraft industry in general, this is a sign that China is continuing to grow in acceptance of civil aviation (as opposed to military or commercial).  This could mean huge growth in the market as the world’s most populous nation gets interested in flying both for business and recreation.

Bell 407GX First Civil Helicopter To Land at Hongqiao | AIN Online

UAVs for Emergency Medicine

A great opinion article is up at EMS1.com about the potential for small and medium sized UAVs in emergency medicine [hat tip to Thomas J. Coyle III on DIY Drones]. The basic premise that the paramedics can use a UAV to protect themselves from a dangerous scene by remotely reconnoitering, and save time by triaging patients before they even arrive on scene.

Conceptual image of an EMS UAV.  Aircraft shown is a Hornet Micro built by Adaptive Flight, Inc.
Conceptual image of an EMS UAV. Aircraft shown is a Hornet Micro built by Adaptive Flight, Inc.

It reinforces two points that I’ve already made before (here and here).  First, that UAVs are an inherently neutral tool–they can be used for good or for evil.  Second, the grassroots–end users–are coming up with ways to exploit the UAVs strengths and benefits faster than our lumbering government can come up with rules.

LiveLoad Bridge Classification App

When I was a young engineer officer, we were introduced to the GTA 05-07-13 method of classifying bridges according to the load capacity. The method was algorithmic, and even though it was the “rapid” field classification method, it could take 4-5 minutes even for a proficient officer with an electronic calculator.  If the individual were required to do arithmetic with a slide rule or by hand, it could easily take 10 to 15 minutes.

That seemed ridiculous to me in the digital age, when laptop computers were everywhere, and we had FBCB2 computers in every vehicle in 4th Infantry Division, that there wasn’t an application or something that could automate this mechanical process.  Not knowing anything about programming and having other fish to fry in my first unit, I tucked it all in the back of my brain until I got to graduate school.  Suddenly, I knew how to program; and, needing practice, I started creating an application on the side that would perform bridge calculations as an exercise.

Well, here we are four years later and I’ve basically finished working on it, and want to make it available to the world. Using this application, a proficient operator could reduce computation time to about 30 seconds–just enough time to type in the parameters and click ‘Calculate’.

I’m making it available for free, and open sourcing the code on SourceForge so anyone can see how it works. Download it from SourceForge now.

Any Sappers, Seabees, Red Horse, USMC engineers, or even engineers our English-speaking allies–if you’re interested in software with even more capability, contact me.  I have the capability to code up calculators for:

  • bridge classification using the analytical method or correlation of curves
  • nonstandard military fixed bridge design
  • vehicle MLC rating
  • concrete formwork
  • culvert and open channel design
  • support bridges (MGB/HDSB)
  • panel bridges (Bailey, Mabey-Johnson)
  • assault float bridges
  • demolitions,
  • minefields/obstacles,
  • dozer/ACE blade teams,
  • flexible pavement design,
  • concrete mix design
  • and basically any calculation in the engineer field manuals that uses step-by-step computation.

Additionally, while it would take a little bit longer, I’m capable of converting this app (or any of the above) into an Android or iOS app.

Screenshots below:

Default ViewConcrete Deck Page

Masonry Arch TabArch Factors Dialog Box


10 Things I Learned About Simulation From GUST, Recap

Over the last two weeks, I’ve put together a string of articles detailing the lessons that I’ve learned over the past four years, first writing my own simple simulations, then experiencing the far more advanced Georgia Tech UAV Simulation tool.  Here is a rundown of the series, with links to each article:

  1. It is critically important to know the ins and outs of software and how computers work.
  2. Good documentation of code is really important.
  3. You have to collaborate with others on all but the simplest code and there are many tools that make collaboration more efficient and effective.
  4. Modularity is critical and must be integrated into the simulation from the ground up.
  5. Spending the extra time on visualization is usually worth the effort.
  6. A good simulation needs the right level of fidelity, which is usually a little bit more than the level you think it needs.
  7. It is extremely valuable to be able to tinker with parameters while the simulation is running.
  8. The simulator must constantly be validated against available empirical data, and updated to fix discrepancies.
  9. Running an experiment in simulation requires the same diligence, forethought, and organization that takes to design a real-life experiment.
  10. Every simulation needs integrated data recording from the very start.


Courtesy Club UMADD.

10 Things I Learned About Simulation From GUST, Part X

Every simulation needs integrated data recording from the very start.

In some of the early simulations I put together, I would run a test case and output the states that I wanted to look at. Then, inevitably, I would realize that I needed to see some other data–maybe the computed forces or the commands generated by the guidance logic. So I’d have to rework the data recording code. Then often the same thing would happen again. Sigh. If only I had planned a little better and thought about the simulation as an experimental apparatus.

GUST, on the other hand has data recording facilities out the wazoo. It records by default all of the data shared over the wireless datalink and all of the inputs and outputs of the console. With a single command, all of the raw sensor data and the outputs of the navigation filter are recorded too. With a little more effort, every single variable accessible in the variable browser is recordable. The graphics output can be saved as a sequence of images, also with a single command. It is clear that data collection has been central to development of GUST from the very start.


A couple of tips for setting up good data collection: Make sure that time and date are included in the data recording, perhaps in the file name itself so that you can find the data a year or two later when you need to publish it, and design the data file to be parseable by Matlab or Excel.

Getting down in the weeds (automatically) . . .

Pretty cool deployment of a quadrotor in “precision agriculture.” The quadrotor represents a cheaper, more precise and more effective platform for gathering data on weed formation than the higher-altitude manned aircraft currently in use for this purpose.

Higher resolution spectral images = more precise targeting of weeds with herbicide = cheaper agriculture and less environmental impact.

The next evolution of this will be UAVs which automatically deploy, decide where they need to look, plan their own flight path, land and charge themselves while sending a set of instructions to the robot responsible for weed spraying.  Kind of like a flying agricultural Roomba.

PLOS ONE: Configuration and Specifications of an Unmanned Aerial Vehicle (UAV) for Early Site Specific Weed Management