• RHETT ALLAIN SCIENCE
•  
• DATE OF PUBLICATION: 07.20.15.07.20.15
•  
• TIME OF PUBLICATION: 10:00 AM.10:00 AM

UH OH: THAT VIDEO OF A PISTOL-FIRING DRONE IS PROBABLY REAL

http://www.wired.com/2015/07/uh-oh-video-pistol-firing-drone-probab...

LET ME FIRST say that this isn’t only a bad idea, it’s most likely illegal. Ok, but is it real or fake? That’s why I’m here. Really, I’m like a fireman in the fire station. I mostly sit around doing things like washing the fire truck and entertaining school kids. However, every once in a while there is an actual internet emergency. So let’s get to work. 

Recoil Physics

I know everyone is in a hurry, but a quick look at the momentum principle won’t hurt. The momentum principle states that the net force on an object changes its momentum.

Now with the momentum principle, we can consider a bullet fired from a gun.

While firing, the gun pushes on the bullet and the bullet pushes back on the gun. Actually, these are just two ends of just one interaction (bullet interacting with the gun) such that the magnitudes of these forces are the same. Further more, the time that the gun pushes on the bullet is the same time interval that the bullet pushes on the gun. If I look at the change in momentum for the bullet, I get:

If I wrote a similar equation for the change in momentum of the gun, it would have the force of the bullet on the gun instead. However, this is the SAME force (but in the opposite direction). This means that the change in momentum of the bullet is the opposite of the change in momentum of the gun. We could also say that the total momentum before the bullet is fired must be the same as the total momentum just after the bullet is fired. This is called conservation of momentum.

Now we can look at the bullet fired from the gun. Let me go ahead and call the whole drone the “gun” just for simplicity. Also, since the bullet is fired horizontally, I can just deal with momentum in the horizontal direction (so it won’t be a vector). Finally, both the bullet and gun start at rest so that I can say the initial total momentum is zero.

Since the initial momentum is zero, the final momentum is zero. This means that I can solve for the mass of the gun. Why solve for the mass of the gun? Well, I should be able to estimate the mass and velocity of the bullet and I can get the recoil velocity of the gun from the video. Solving for the drone-gun mass will be one method to check the reasonableness (is that word) of the video.

But wait! Is momentum always conserved? No. Momentum is only conserved for a system with no net external forces. For the case of the drone-gun, this is mostly true. After the bullet is fired there can be some thrust forces to slow the recoil. However, in the instant right after the bullet is fired it should be a fairly nice approximation to say that momentum is conserved.

Video Analysis

Now I just need to load this video into my favorite (and free) video analysis software—Tracker Video Analysis. Really, the only thing I need to find the velocity is something to scale the video. Fortunately in the description of the video it says that the drone is 26 inches from the muzzle to the rear of the frame. Oh, I also have to account for the background motion due to a moving camera—but that’s not very difficult.

Looking at one of the shots, here is the horizontal position of the drone-gun.

By fitting a linear function to the position data, I can get the recoil velocity. From this graph, you can see that the drone-gun has a recoil speed of 0.475 m/s. Looking at another shot, I get a recoil speed of 0.65 m/s. Let’s just say the recoil speed is about 0.5 m/s (as an estimate).

Now, what about the bullet? I am just going to wildly guess that this is a 9mm bullet. The last time I looked at a 9mm bullet, I said it had a mass of 7.45 grams and a speed of 435 m/s. Putting these values (along with the recoil speed of the drone-gun), I get a drone plus gun mass of 6.48 kg or 14.2 pounds. That seems a bit heavy to me. Suppose the drone is firing a .22 caliber bullet with a mass of 2 grams and a speed of 300 m/s. This would put the drone-gun mass at just 1.2 kg. If the pistol has a mass of 0.5 kg, the rest of the drone would be just 0.7 kg. That seems fairly light for a drone—but if the bullet mass is higher then the drone mass is also higher. Finally, if the pistol fires a 9mm round it could be going much slower—even around 300 m/s. This lower bullet speed would put the drone mass at 4.47 kg.

Looking For Fake Camera Shake

There are several ways you can spot a fake video—here is a description of my favorite tests. One of these tests looks at the background motion for a moving camera. If you want to add special effects to a movie, one method takes a video recorded on a tripod so that it’s easier to add the effects. Of course after that you want to make it look like it’s a hand-held camera so you add “fake shake”. However, it turns out that fake shake is detectable. Here is the motion of the background for the drone-gun video along with a known fake video.

The drone-gun video does not look like it has fake shake.

Conclusion

Sadly, I think this video is real. Why is this sad? Because once you start putting weapons on drones, it’s just a small step towards robots taking over the world.

---

Gun-firing 'drone' built by US teen under investigation

http://www.msn.com/en-us/news/us/gun-firing-drone-built-by-us-teen-...

© Provided by AFP The Federal Aviation Administration is investigating an online video that shows an alleged home-made "drone" firing a handgun in the Connecticut countryside

The Federal Aviation Administration said Tuesday it was investigating an online video that shows an alleged home-made "drone" firing a handgun in the Connecticut countryside.

The 14-second video called "Flying Gun" shows a homemade multi-rotor hovering off the ground, buzzing furiously and firing a semiautomatic handgun four times at an unseen target. 

It was posted on YouTube on July 10 and has been watched nearly two million times. It was filmed by 18-year-old Austin Haughwout from Clinton, Connecticut.

Haughwout is studying for a degree in mechanical engineering. Neither he nor police could be reached for immediate comment.

"The FAA will investigate the operation of an unmanned aircraft system in a Connecticut park to determine if any Federal Aviation Regulations were violated," it said in a statement.

"The FAA will also work with its law enforcement partners to determine if there were any violations of criminal statutes."

But Haughwout's father denied his son had built a drone, which he said are pre-programmed, and said this device was manually controlled.

"People have been playing with RC (remote-controlled) toys for many decades," he told AFP by telephone.

"The proper name for this is an RC quadcopter. The media keeps using the inappropriate word because it helps you to generate fear," he added.

Haughwout said the FAA had made no contact with his family.

"I don't understand why people are making such a big deal of it. It's not like it's anything new," he said.

"He's a mechanical engineering student. He builds all different kind of things."