When drones and pork collide

When drones and pork collide
A pork roast meets a drone propellor as part of safety research
A pork roast meets a drone propellor as part of safety research
View 1 Image
A pork roast meets a drone propellor as part of safety research
A pork roast meets a drone propellor as part of safety research

In a display of true awesomeness, researchers from the Drone Research Lab at Denmark's Aalborg University have built a catapult that fires hobby drones at a pork roast. This display of porcine projectile piercing is not only high on the list of conversation starters, it also has the serious goal of finding out more about the safety hazards the tiny aircraft pose to people and property.

Anyone who has given their kids a toy drone knows that they can and will crash into just about anything. The problem is, how much damage potential do various hobby drones pose and what kind of injuries can they inflict? To work this out, the Drone Lab and the Aalborg University Hospital came up with a motorized catapult and high-speed camera system that can make precise measurements of both the speed and the force of drone impacts.

Despite its name, the catapult doesn't hurl drones across the room like bean bags. Instead, the drones or drone parts are attached to a slide and are pulled along an aluminum rail by an electric motor until they slam into a variety of targets at the other end, such as a pork roast standing in for an innocent passerby.

According to the team, the 3 meter-long (9.8 ft) catapult shoots a 1-kg (2.2-lb) drone at 15 m/sec (49 ft/sec) while the camera records the mayhem at 3,000 frames per second. Meanwhile, the force of impact is measured over time to ascertain the severity of the injuries.

The rationale behind the Aalborg catapult is the same as most modern safety testing and goes back to the early days of aircraft safety. It may seem strange to shoot drones at pork, but in 1943 Westinghouse Electric started shooting dead chickens at aircraft at 200 mph (322 km/h) using a compressed air cannon (a technique that was used well into the jet age), to test the effects of bird impacts on windscreens and engine intakes.

The researchers say that the catapult is still in the early stages of development as the team works to refine the mechanics and electronics, as well as the camera and lighting setup. As the experiment progresses, the catapult will be tweaked to fling larger drones at higher speeds.

The top video below shows the catapult taking on a pork roast in slow motion, while the one below shows the same impact at normal speed.

Source: Aalborg University

DroneImpact - T0005 - 3200 fps to 25 fps (Pork, carbon propeller, 9 m/s),

DroneImpact - T0005 - 25 fps (Pork, carbon propeller, 9 m/s)

Mel Tisdale
Surely the results to date are sufficient to show that drone rotor blades need to have cowlings that can act as a guard. Or do a set number of serious injuries have to occur before the legislators decide to act? Elsewhere in today's issue there is an article about drone racing. If it is going to be successful, it will need to be in close proximity to the spectators. If any drones need guarded rotors it has to be those taking part in this sport.
I very strongly doubt that some drone usually available in the market can achieve speeds like these, also, it won't have the kind of momentum they had in this setup. So basically the momentum did what it did to the pork not the harmless rotor. So a waste of effort and absolutely misleading.
So an animal's life is taken so it's meat can be subjected to such a disrespected use.
This is a very bad test setup. The rotor is mounted to a rigid sled with a bearing system that can impart much higher forces to the blade than what could ever be created in an actual collision. The motor prop assembly should be mounted on a weighted structure that simulates the mass of actual drones and carried in a sled that stops and disengages just before the blade makes contact. That way the reactive forces from the impact with the test target will be able to affect the motion of the drone the way it would in a real impact. Those reactive forces will cause the drone to deflect and rotate away from the target which it cannot do because it's mounted to the rigid structure of the sled. I have not seen a single report of anyone being seriously hurt by one of these small light weight devices. I have seen many reports and videos of people being seriously injured and sometimes killed while watching motor sports, golf, hockey and baseball. So let's test this the right way and keep things in perspective.
Nathan Holmes
@UmairBinAsim I was also skeptical, but on the momentum front this is actually pretty solid. The phantom 3 is available for around half a grand, so seems a legitimate litmus test. According to its specs it has a mass of 1.2kg compared to the 1.5kg of the dolly in the test, but a maximum horizontal velocity of 16m/s whereas the test is quoted at 9m/s, giving the drone a higher potential momentum than this test. There are still issues with freedom of movement, replication rate and carbon vs plastic, but I'm a bit more convinced than I was at first impression.
I take great offense at being simulated by a hunk of pork. No volunteers? Where are the graduate students?
Another scare tactic misleading demonstration staged to allow the viewer to come away with a false sense of doom as it relates to drones. This is so unbalanced and biased, it's a wonder any respectable university allows its name associated with these results, even though no conclusions were "officially" made. Let's just look at a few... ..."one of the popular small hobby drones hit objects or people"... WRONG, that is not a "popular small hobby drone" at all. It is a controlled test fixture using carbon fiber propeller that are NOT used across the board with "popular small hobby drone[s] ". The VAST majority of "popular small hobby drone" use styrene, which even in this "Non real world demonstration" would have caused considerably different result. Drones, with minimal practice and skill, excel in their ability to easily hover and be controlled compared to model helicopters. This quality allows them to be used by amateur from almost any starting point. This maneuverability is their strong point. The flyer's goal for the most point is to fly in a deliberate way as to capitalize on this ability, not to race with them as would be the case erroneously depicted here attempts to represent. Again, "popular small hobby drones" cannot even come close to speed represented here without a huge tail wind. Can "popular small hobby drone[s]" race? Of course. But the 33 MPH scenario of this staged situation is far from probable. Real world momentum is an unknown factor in this test. If we formed a hollow hatchet out of aluminum foil and could sling it at you at 33 MPH, vs. slinging a three pound puffy feather pillow at you at 33 MPH, which would cause more damage? The puffy feather pillow would do the most harm. We don't know the power output of what is being used as the acceleration source of the "Test Sled" or if it was still under forward power at impact, or the weight of the "Sled". Things move when impacted which tends to minimize destructive results. That's why cars have and rely on "Crumble Zones". In this "test" the pork is part of the chassis of the test fixture and incapable of movement. That is the equivalent of the pork placed against of a concrete wall and hit with a three pound sledge hammer rather than the piece of pork being swatted and allowed to move across the grass. I can assure you that when the three pound pillow makes contact with you at 33 MPH that you will be knocked down, and if it wasn't for that, your face would receive the same tenderizing that the pork against the concrete would have. There are many other all too obvious misinterpretations that an unrealistic "apples and oranges" comparison like this unrealistic demonstration will unfortunately lead to.
9 M/sec is 20mph. That is top speed for most drones. The vast majority of drone mishaps are drones falling from the sky, not running straight into people at top speed. A further flaw in this test is the angle of attack that the prop is approaching the target at. At 20mph the drone would be tilted forward considerably and likely the blade would have snapped rather than impaling the target. Furthermore, the blade hit perfectly perpendicular to the target which is a low odds impact. I wonder how many tests they had to run before they got the result they are showing now. This is clearly dishonest anti-drone propaganda. Aren't these the same jerks who train raptors to attack drones. I really worry about injury to the bird doing that. I am not saying that these things are harmless but this is ridiculous.
D Jazani
Most people objecting to the test do not understand propellers. I am a modeller of 40+ years and have seen my fair share of propeller injuries, from chopped off fingers to chunks taken out of fore arms (even chopped through overalls). The cage around most drone props is there to protect the propeller against breakage, it will deflect and still allow propeller slicing against objects (Alicia next time there is a drone around try to caress the prop be it styrene, just have a doctor on stand by, enjoy the "Crumble zone" and I hope that you will never be caught in "Crumple zone" a pillow at 30mph can be swung by a 9 year old). The high pitch sound that you are hearing from all the drones is the supersonic sound of the propeller tip. Don't belive me? Umair use a calculator D*Pi*rpm much faster than the speed of sound at about 340m/sec. Those non-believers that object and profess about things being wrongly set-up in the experiment are totally wrong. At universities experiments are carried out after background research and identification of a problem has been completed, the proposal is put forward to further research and ascertain the facts, proposal is reviewed by specialist academics PhDs or industry specialists, once approved, experiment is supervised by specialist PhDs, analysed to draw conclusions and recommend solutions (ban public sales, hopefully and restrict use in unpopulated areas) and peer reviewed by specialist PhDs academics as well as industry experts. Those that object and challenge, do you have PhDs or industrial experience in the field? Calling people names (Corniche) is the sign of poor knowledge and running out of counter arguments. Flying toys (makes no difference on the price of the drone, big boys toys are more expensive) for a few years does not make you an expert, design and build them (not assemble from a kit, or upgrade a kit with better motors, etc). Most people buy RTF and do not even understand the science of flight what they are playing with. I made flying models from scratch, starting with rubber powered, moved into glow-fuel (and yes I used a "chicken stick" and was never injured, because I understood the danger) and then electric brushed and then brushless. I lecture on motors, batteries and mechatronics, so reasonably informed about these things. For the ones that say people have not been hurt (n2liberty), the last one I remember was a little boy maimed for life by an idiotic family friend; http://www.telegraph.co.uk/news/uknews/12022429/Toddler-loses-eye-after-drone-propeller-sliced-through-it.html Testing on pork is as close as testing on human (facile comments of testing on students as just silly), BTW pork is used for forensic analysis because it physically resembles humans in skin, sub-cutaneous fat and muscle structure/density. Rotary wing craft fall out of the sky (it happens to the best, I was a witness to Robbe team accident and many others in air shows), no gliding ability. We have carried out experiments with UAVs and could easily interfere with Tx/Rx, take over control and interfere with UAV, majority RTF UAVs have poor Tx/Rx security. When they do fall Newton's law rules and accelerates them nicely groundwards for destruction and causing destruction as well. I am totally for controlling UAVs and not allowing every financially able person to have one unless they are trained, flying in the designated areas and are fully insured. Just like all licensed RC flying clubs.
amazed W1
Alicia Russell is so right to point out that "apples and oranges" comparisons can be disastrous, true reality has to be modelled as closely as possible, and this is why we make prototypes. Rolls Royce and the British Rail development "laboratories" used to be very close and the RR joke was that they felt they should tell BR not to use frozen chickens when modelling bird strikes on high speed trains.