Ask Hackaday: How Would You Detect A Marauding Drone?

Drone over a wheat field

The last few days have seen drone stories in the news, as London’s Gatwick airport remained closed for a couple of days amid a spate of drone reports. The police remained baffled, arrested a couple who turned out to be blameless, and finally admitted that there was a possibility the drone could not have existed at all. It emerged that a problem with the investigation lay in there being no means to detect a drone beyond the eyesight of people on the ground, and as we have explored in these pages already, eyewitness reports are not always trustworthy.

Not much use against a small and mostly plastic multirotor. Sixflashphoto [CC BY-SA 4.0]
Not much use against a small and mostly plastic multirotor. Sixflashphoto [CC BY-SA 4.0]

Radar Can’t See Them

It seems odd at first sight, that a 21st century airport lacks the ability to spot a drone in the air above it, but a few calls to friends of Hackaday in the business made it clear that drones are extremely difficult to spot using the radar systems on a typical airport. A system designed to track huge metal airliners at significant altitude is not suitable for watching tiny mostly-plastic machines viewed side-on at the low altitudes. We’re told at best an intermittent trace appears, but for the majority of drones there is simply no trace on a radar screen.

We’re sure that some large players in the world of defence radar are queueing up to offer multi-million-dollar systems to airports worldwide, panicked into big spending by the Gatwick story, but idle hackerspace chat on the matter makes us ask the question: Just how difficult would it be to detect a drone in flight over an airport? A quick Google search reveals multiple products purporting to be drone detectors, but wouldn’t airports such as Gatwick already be using them if they were any good? The Hackaday readership never fail to impress us with their ingenuity, so how would you do it?

Can You Hear What You Can’t See?

It’s easy to pose that question as a Hackaday scribe, so to get the ball rolling here’s my first thought on how I’d do it. I always hear a multirotor and look up to see it, so I’d take the approach of listening for the distinctive sound of multirotor propellers. Could the auditory signature of high-RPM brushless motors be detected amidst the roar of sound near airports?

I’m imagining a network of Rasberry Pi boards each with a microphone attached, doing some real-time audio spectrum analysis to spot the likely frequency signature of the drone. Of course it’s easy to just say that as a hardware person with a background in the publishing business, so would a software specialist take that tack too? Or would you go for a radar approach, or perhaps even an infra-red one? Could you sense the heat signature of a multirotor, as their parts become quite hot in flight?

Whatever you think might work as a drone detection system, give it a spin in the comments. We’d suggest that people have the confidence to build something, and maybe even enter it in the Hackaday Prize when the time comes around. Come on, what have you got to lose!

117 thoughts on “Ask Hackaday: How Would You Detect A Marauding Drone?

  1. if it’s a “drone” with a data link back to a mobile device, they present as an AP

    just look for a new access point showing up
    for a home built, look for 2.4Ghz, 5.8Ghz or 900Mhz video

    1. This is purely speculation but I imagine the fact they haven’t caught the perpetrators at Gatwick means the drone(s) employed were either flying to predetermined co-ordinates using GPS or made use of wireless mobile data networks for control.

      1. riiiight….

        that assumes there was a drone

        so your saying it was a home built drone, hell bent on an act of terrorist destruction???

        which drone can you buy that will let you program a course??

        flying anything with a mobile network is going to be really hard, if not impossible

        I’ve tried driving a UGV with 3G, it was only possible at a crawl

        to fly something over a mobile network is a whole “next level”

        say it with me

        “there was no drone”

        1. I have flown a drone on a public 4G network. The control latency (downlink is about 50-100ms). No good for stabilize/acro modes, but good enough for GPS-assisted velocity/position control mode.

        2. It’s possible to remotely direct craft without real time control (see: New Horizons, they’re not exactly steering that thing directly). There’s quite a few prosumer drones that can follow GPS waypoints.

          As for terrorism, there’s plenty of room between upstanding citizen quadcopter pilot and terrorist bent on destruction, including prankster with a disregard for the safety of others.

        3. Most of the ‘sightings’ were of the Police’s drone. Not much point in taking pictures of their own drone as it would make them look even more amature, if that was possible….

        4. “which drone can you buy that will let you program a course??” My $90 quad can be programmed with a set of waypoints to fly autonomously, I believe the Parrot quad and fixed wing drones can all be programmed with waypoints.

      1. in controlled airspace?
        airports are like rooly, rooly big

        I’ve been flying RC aircraft since the 70’s, one of the reasons I hate friggen drones so much

        we started with 27Mhz, went to 37Mhz, then 2.4Ghz

        every “improvement” to RC transmitters has been to limit range

        what “background noise” are you talking about???

        just employ some folks to look in the ISM bands

        this is a beat up, plain and simple

    2. You have a good point, though it’s not as simple as looking for a wifi AP. A drone might be using LTE or any number of other C&C technologies.

      How about two or more directional antennas spaced around the perimeter of the protected space? Any unaccounted for transmitter within the protected space would be suspect.

    3. That’s not always the case. You are assuming the operator is using an older dji drone or a toy drone, both of which won’t be detected from the air over an airport unless they are using a high gain directional wifi antenna. There is no device that can alert you to a drone nearby. As the article said, they would already be using them if they worked. The cost of shutting down an airport fast exceeds the cost to purchase commercial equipment. The only real way to detect a drone is with vision. They need eyes.

      1. I don’t see a problem with using radar – sure it needs to be on a different frequency to anything in use, and would probably need a large number of small antennae all round the field rather than one great big one, but you don’t need the extreme range of the big boy either.

        1. Radar is wonderful, but the issue here with radar is its lower resolution at radio frequencies compared to higher frequencies (like infrared or optical bands) needed to detect smaller objects. You can increase the frequency of the radar to improve resolution (and reduce range due to the atmosphere weakening the signal) but the hardware gets significantly more advanced and expensive as you climb higher into the microwave bands :P

  2. Most are radio transmitters. Use standard triangulation techniques.

    You can also use the motion detection hardware present in most security camera chips. Here’s a cheap board that is perfect for the job. About $130 including two 4K cameras.

    http://www.lindeni.org/lindenis-v5.html

    Dual 4K cameras with hardware binocular depth perception. Hardware motion detection. A visual AI engine that can look for drones in the areas flagged as having motion. Teach it to differentiate drone from bird. Ability to relay the flagged image to a human for verification. At $130 you can afford to put a 1,000 of these around an airport aimed all over the place.

    This even has a chance of working in the dark. You can take the IR filters off from the cameras. The drone should be warmer than the sky background. Plus the motion detection will detect anything, if the drone is move in front of distant lights, it will still be detected as motion even if it is not visible.

    Here is sound tracking software for array microphones. Check out the pictures…
    https://github.com/introlab/odas
    It might be possible to get auditory tracks off from the engine whine. v5 has support for 8 channel mics. And you can feed it into the hardware AI engine.

    Combining the angle/depth info from the cameras with tracks from the auditory system and human gatekeeping, it may be possible to develop a guided takedown interceptor. Note that the camera hardware is certainly light enough to put on your own drones to provide better tracking at terminal interception.

    1. Someone hasn’t been involved with development, support and maintenance of complex real-world H&S mission critical systems … 1000 cameras running on low cost hobbiest grade hardware??? Yeah right.

    2. “Teach it to differentiate drone from bird. ” Thinks like this can quite often be the gotcha. This would need to be very good and provide a low level of force positives with practically zero false negatives.

      1. AI vision system are actually pretty good at this. But the system is not going to be 100% automatic. When the vision system reaches 80-90% confidence that it is seeing something drone-like it is going to alert a human to look at the camera feed.

        This is also not a one-shot deal. The vision system will capture many frames with the object of interest from varying angles. One of those angle may trigger a 100% certain drone detection event from the vision system.

    3. The claims made for image recognition are universally way beyond the practical reality – that will either detect nothing or it will flag everything from wind-blown leaves to Dreamliners.

      For very narrow cases (ANPR, left-object, counting, tripwire / line crossing), video analytics is getting as far as being pretty okay… but picking a drone out of the sky over an airport is not going to be easy.

      Also, two 4K cameras for $100 suggests low quality and poor lens so you’re going to struggle – there are plenty of cheap 1080p cameras out there which provide worse picture quality than the older professional PAL/NTSC units.

      1. It’s dev board, not a production system. The image sensors are on connectors so can can replace them with $200 Sony starlight sensors if you want. The lens mounts are interchangeable too. The idea is to prototype a working system, later you can make production hardware in weather proof packaging.

  3. The RADAR systems at airports already have the technology to detect UAS and many companies make off the shelf RADAR solutions that can be deployed anywhere. The RADAR cross section of a UAS is almost identical to that of a bird and airports have had bird RADAR for decades, the system just needs a software update to process the new data. However this is both expensive and, at least in the US, the threat of UAS is not being taken seriously.

    RADAR and RF UAS detectors have been in use for major events for years including NASCAR, The Superbowl, etc. The agencies deploying this technology just aren’t advertising.

          1. WOW!

            If I saw only the X27 clips from that video, without context, I might think they were shot in broad daylight. The quality is amazing. It looks like if you have one of those cameras, there’s basically no such thing as night anymore.

            I wonder if there’s any publicly known info on how it works…

  4. Haven’t they spent £5m on milimeterwave radar and long range thermal? That’s what was installed on the roof in the news report photos.

    The military can spot, track and neutralise something as small as a falling mortar shell. A drone should be a piece of cake.

    The best ‘hackaday project’ solution would be a team of hunter drones flying a perimeter circuit, spaced 20m apart. After each lap they land and recharge. At the same time, another takes off.

    1. Not all cell phones are created equal. The Samsung S9+ has 10X zoom but it is hard to record something that small.

      Mr York said: “It is established that the incident, which led to widespread travel disruption, was caused by numerous instances of illegal drone activity at the site between 19 and 21 December. This was not a police drone.

      “Police, supported by a range of partner agencies and working closely with Gatwick Airport, are currently examining relevant sightings by 115 witnesses – 93 from credible witnesses, people used to working in the busy airport environment, including a pilot, airport staff and police officers.

      https://news.sky.com/story/some-gatwick-drone-sightings-may-have-been-police-drones-chief-constable-says-11593854

      Police later backtracked from saying there was no drone:

      More than 200 reports of drone sightings were made and police took 67 statements, including from police officers and airport workers.

      In a statement, Gatwick said “there were multiple confirmed sightings of drone activity at the airport”.

      https://news.sky.com/story/gatwick-drones-police-communication-was-a-mess-up-say-ministers-11590916

      Safety tops the list of urgent issues, especially in the wake of two international incidents last month. The Mexican airline Grupo Aeromexico SAB is still investigating whether a drone crashed into a Boeing 737 jet as the aircraft approached a runway in Tijuana on Dec. 12. The jetliner landed without further incident, but its nose was crushed and partially sheared off in the collision.

      https://www.washingtontimes.com/news/2019/jan/1/drones-likely-to-be-regulated-by-state-city-govern/

  5. Wait for a problem to happen and go from there. Seriously. I’m not convinced this is a legitimate threat, it’s just technophobia and sensationalism and fear culture. We obviously have anxiety issues concerning airports. The security creep is never going to be enough, and the ridiculous amount of precautions we already have are based on very little evidence. Enough already.

    Don’t worry about the toy RC helicopters which have been around far longer than people have been building quads and calling them “drones.” Never has caused a significant issue. There are better things to spend this energy on.

      1. Not so easy. The minimum output power for ADS-B is something on the order of 60W (forgot the exact number). This may be challenging to do on a small drone, but more importantly, many transmitters at this power level will saturate the network. Many more ground stations would need to be added (or the minimum power lowered for small drones, but that is a very slow process)…

      1. Non-compliance could be make harder by requiring manufacturers to build them into the drone. It could be designed in such a way as to make it difficult to remove without impairing the structural integrity.

      2. ah the old “criminals don’t follow laws so why have laws at all” argument, the hallmark of the intellectually dishonest poster who’s only interest in muddling the water so nothing gets done. translated into none douche bag its read as ” who cares about the rest of you useless trash, i don’t care if you all die as long as i can do what i want.”

  6. Any commercial drone will transmit on one of the few, narrow open bands (ie, 433MHz, 868Mhz, 2400MHz, 5GHz etc). Usually omnidirectional. This should be enough to use passive radio antennas to locate them.

  7. Most are radio transmitters. Use standard triangulation techniques.

    You can also use the motion detection hardware present in most security camera chips. Here’s a cheap board that is perfect for the job. About $130 including two 4K cameras.

    www. lindeni. org/lindenis-v5.html

    Dual 4K cameras with hardware binocular depth perception. Hardware motion detection. A visual AI engine that can look for drones in the areas flagged as having motion. Teach it to differentiate drone from bird. Ability to relay the flagged image to a human for verification. At $130 you can afford to put a 1,000 of these around an airport aimed all over the place.

    This even has a chance of working in the dark. You can take the IR filters off from the cameras. The drone should be warmer than the sky background. Plus the motion detection will detect anything, if the drone is move in front of distant lights, it will still be detected as motion even if it is not visible.

    Here is sound tracking software for array microphones. Check out the pictures…
    github. com/introlab/odas
    It might be possible to get auditory tracks off from the engine whine. v5 has support for 8 channel mics. And you can feed it into the hardware AI engine.

    Combining the angle/depth info from the cameras with tracks from the auditory system and human gatekeeping, it may be possible to develop a guided takedown interceptor. Note that the camera hardware is certainly light enough to put on your own drones to provide better tracking at terminal interception.

    1. “This even has a chance of working in the dark. You can take the IR filters off from the cameras. The drone should be warmer than the sky background.” Those 4k cameras are only for visible light and near IR, and are incapable of thermal imaging. Even if you use a real thermal camera, it would still be kind of dodgy. Depending on what thermal sensor you use, it would be incredibly difficult to actually spot a drone at any meaningful distance with anything uncooled or even remotely cheap.

        1. That’s a good one. IR searchlights are also a cool possibility. IR doesn’t have the same reflection properties as normal light does, which could be a bonus or a downside. The IR laser grid would also offer location tracking and rangefinding, because the computer could know where the laser was pointed when it received the reflection, and there could be coded laser pulses like in the laser speed guns.

  8. Microphone arrays won’t work as you’re going to need to spread them all over the place since at a distance it’s really difficult to isolate rotor sound from background noise, especially if it’s windy.
    Radars won’t work either as small drones cross section is not that different from birds, moreover with a target flying so low reflections from the ground, especially if wet, will make it almost impossible to spot clearly objects so small.
    Cameras have hard time working consistently with variable lighting conditions and thermal cameras are usually not resolute enough to reliably spot hear from rotors at a distance.
    Spectrum analisys works well as long as someone is transmitting something. As some pointed out, a serious attacker would set up waypoints and disable any radio communication so that the drone would fly itself to the target using GPS+IMU+altimeter and flight controller to stabilize flight and drop whatever payload at target arrival.
    I’ve contributed to developing a RF spectrum surveyor that is able to demodulate and decode both control and telemetry and in most cases we can not only determine presence but also intercept telemetry, take control or jam communication so that drone without control would likely land.
    This is commercially sold in conjunction with radar, direction finder, cameras and even mic arrays but in the end the only effective system is rf analisys and of course only if drone is transmitting telemetry or receiving commands…

  9. “RADAR CAN’T SEE THEM”
    WRONG
    crappy centimeter-wave radar that technologically belongs into the 70s can’t see them, I assure you that modern milimeter-wave radars (especially with phased array antennas), like those used for small anti-missile systems could pick them up easy.

    Also, the drone has radio emissions that are not normally present in that space, so you could literally grab one of the several relatievely cheap (compared to an active milimeter-wave AESA radar) off-the-shelf passive radar solutions, install it and have a bit of code added, so that it will automatically flag a possible drone, track it in 3D space and do the same for the remote control.
    Once you can track the thing in 3D space, you can deal with it by multiple methods.

    1. Actually I didn’t write radar can’t see drones but rather it would be useless because drones can’t be easily differentiated from birds, moreover since both fly low it’s difficult to avoid ground reflections from people, cars, etc, especially at distances of a few km.

      1. Birds don’t move in predictable patterns, drones mostly do.
        Most birds don’t hover for any extended period of time. Yes, there are exceptions, but those could be dealt with by DSP-ing the radar signature (flapping wings).

        BTW a big AESA radar is potential HERF weapon itself, so it can “interrogate” potential threats just by using a much longer pulse.
        Fleshy intruders would be fine, drones on the other hand would almost universally suffer from interference the microwaves would induce in nearly everything (not just the flight controller, but even ESCs). If not immediate damage to the electronics, just crashing from loosing stability is still a good result.

  10. Not a technical solution but, if there was an active threat such as at Gatwick. I’d ask the local plane spotters to cover one half of the perimeter, and the local twitchers ( bird watchers ) to do the other.

    You could give a bounty on spotting and photographing a drone but I suspect people would do it for kudos. Spotters and twitchers tend to have good kit, and know how to use it. I suspect the police have very few photographers trained as well as a decent spotter and the police force’s budget may not stretch to the toys that a dedicated and fairly well paid plane spotter will acquire.

    Spotters are also likely to have more knowledge of what a drone looks like than the average policeman / pilot.

  11. So, what would you do if someone released a few helium filled party balloons with pieces of tin foil attached to them upwind of an airport? Or maybe started tossing ball bearings with a catapult in the direction of a runway? Or… whatever. there’s really thousands of ways to disrupt or endanger flight operations, and regulation will do little to stop them, You can only hope to catch the miscreant, or try and prevent this sort of thing in case it is an organized effort. One thing for sure, if the “idiot of the day” is looking for exposure, publicizing these accidents only gives him and his emulators one more reason to act on his idiotic ideas, while damaging everyone else that happens to share his hobby. Oh, and I do not own a drone.

    1. exactly!

      “the man” having to spend money on “drone prevention” is a sure fire way for more rules, more regulation on our RC aircraft hobby

      everything was fine until those crappy planes and quadcopters started getting sold in supermarkets!!

  12. How about an array of cameras around the airport with infra-red search lights and computerized object recognition? …… Just throwing this in as a random brainstorming idea.

    Ok …. wont work in the rain ….. but what else?

  13. I think a defined scope helps manage expectations. As many here have shown through numerous examples having a perfect solution that works in all environmental conditions and threat environments is going to be difficult. Coming up with a solution against a determined adversary is going to be darn near impossible in most situations. How are you going to be able to counter a threat from a below rooftop, ultra fast drone using non-standard frequencies/spread spectrum, or fully autonomous with no c/c link? Just not really feasible in an urban environment.

    But it doesn’t have to be perfect to be enormously useful to airports, concert venues and other sites where the main threat is from people who buy drones at Amazon and ignore the rules. Seems like a solution involving RF detection/DF along with optical should be able to handle 90% of the annoyance intrusions.

  14. Dear Google, here’s an AI challenge. Flying object detection, identification and tracking using active and passive detectors in a fixed volume. Methods can include but are not limited to PIR, audio analysis, lidar, sonar, radar and em spectrum analysis.

    1. Seriously? So you detect a drone when it’s within 200 meters… How much time do you have left to do something before it’s over your head?
      Guys, lidar, it and so on are not a solution if you want to have an early warning. For who was talking about cameras and object detection maybe some math helps… With a 300mm telephoto lens you have a viewing angle of around 6 degrees. At 1km a 100m object would fill the picture, which means that at full HD a pixel amounts to 5cm so a 50 cm drone would be 10 pixels big… Not much to do any decent image recognition that could give good confidence against birds. Note also that 1km is not that far and that 6 degrees field of view is really small so you’re going to have lots of cameras (doc is limited also vertically!) Or some very complex system to locate and track objects

    1. there was a defcon talk a while ago about knocking drones out of the air using Denial Of Service attacks. that may be a good way to deal with illegal drones, but you still have to detect them first which isn’t easy. plus, just spamming an area with strong radio signals would be considered jamming which i know is illegal for the general public. but i don’t know how those rules apply if you own the area you are jamming and if you can guarantee that you won;t be jamming outside your radius.
      But even with that aside, i don’t know what effect blasting strong radio signals around an airport would have on aircraft instruments, along with the fact that you would probably also be jamming communications between the control tower and individual aircraft.

  15. A bit of a “wild” card but I believe that what’s needed is a mechanical eagle. Something with the flying speed, agility, vision and honing skills of the drone-intercepting bald eagles that the Dutch police trained, but which can – crucially – be manufactured at scale.
    In a quest to find such a solution (or other brilliant ideas) the Delft University of Technology in the Netherlands has set up a spectacular competition called DroneClash – a kind of Robot Wars for drones with €50,000 in prizes up for grabs: https://www.youtube.com/watch?v=R3QXSDELH-A
    It would be fantastic to have some Hackaday ideas put to the test in this arena! Enter by 15th January!

  16. Vision systems are okay but would have a rather difficult time seeing through fog, storms, etc. and high altitudes. my brother flies drones far lower than 400 ft faa regulation and most times you can barely see them due to clouds or overcast. I like vision systems but once you bring them outdoors in the elements they start to lose their appeal.

    Radar dsp could hypothetically filter out birds and other clutter spectra. Radar yields range, velocity, and angle to target and can ‘see’ through various lighting and weather conditions. Assuming birds have typical velocities lower than the average drone, you can filter out drones from other moving clutter spectra, as long as it can be detected above the noise thresholds. Noise is random so it filters itself out! Modern radar is the most robust option imho.

    Texas Instruments sells a new pmcw radar evm. They can even mux Radar info with vision systems.

    1. Because optical flow based navigation etc. is a thing. It would be dangerous to build a protection system that made assumptions based on what is currently available to the consumer mass market, especially given how fast things move in the tech area and the amount of experimentation by hobbyists.

        1. Yes, but it is not so easy. Many drones can fly a mission purely on GNSS navigation. You could try to jam that as well but then you cause significant collateral damage to other systems, especially if you need to use omni-directional antennas since you don’t know where the drone is. Depending on how the drone/receiver is constructed the jammer may need to be above the drone as well, leading to an even greater GPS-denied area.

          And even then, you would need to jam almost the entire spectrum since using a frequency without a license is a much lesser crime in most places than interfering with the safety of aircraft. There are many open source RC systems where you can change the frequency easily. There is also the option to use 2/3/4G to control the device, so you will need to jam mobile phones as well.

          This is something that could work in the middle of nowhere…

    2. That would also jam all cell phones and WiFi at distances greater than 10 miles. And do nothing for autonomous flight. If you also take out GPS you will disable some systems on aircraft and make life suck for people returning rental cars when their nav systems no longer work just when they are needing the best directions.

  17. Hello, I’m not very savy on the Tech talk and words that are used , So I apologise for my lack of.
    There has been a Drone that has been flying around my house for about 2 months now. Also like to add that I love anything that is in reference to Technology. Anyhoo,
    So I researched drones in my area with local law enforcement. It sent me to the FAA website. It said you have to have a UAS pilots license and get a waiver from the FAA to fly in certain Zones. It also sent me to Data. Gov where it shows you what it is, who is the pilot and it will show you what Drones are flying in your area.
    So, I just thought I would share this with you.
    Thanks for letting me Comment.
    Sincerely. A 51 year old white female. ????

  18. After reading año the comments, I’m going to suggest something crazy from the ignorance’s viewpoint:

    The vast majority of small flying devices uses brushless motors, controlled by some sort of ESC. That kind of equipment will probably emit some EMF noise in easily identifiable low frequencies, directly related to the RPM of the motors.

    With some software defined radios and different directional antennas, would be easy to spot and track a drone, even if there isn’t a C&C wireless link.

    1. If we’re talking about possible terrorism/ military drones, they often use gas engines that have nothing to do with ESC’s. (to be fair, I think most people in the replies missed this possibility).

      1. Yes, the gas engines would be undetectable with this method, but they also emit more heat, so would be possible to use a dual system involving FLIR cameras.

        Anyhow, this is like chasing mosquitoes with a laser gun

    2. I have measured this to ensure little/no noise was being emitted in the bands where the radio receivers in the UAV worked on. The interference is extremely weak since the ESC is not coupled to a good antenna. It does reduce the performance of the UAV radios in some cases. The worst result was -65dBm at ~950MHz using the on-drone RX antenna, this is enough to significantly degrade performance, but unlikely to be detectable from any distance.

      1. Maybe the harmonics in the MHz range are weak, but would interesting to study the emission in the KHz bands. I’m pretty sure that the pulsed coils are able to produce quite powerful signals (unless they are perfectly isolated with some sort of Faraday cage).

  19. Someone hasn’t been involved with development, support and maintenance of complex real-world H&S mission critical systems … 1000 cameras running on low cost hobbiest grade hardware??? Yeah right.

  20. What about building a wall around the airport area, instead of just a fence. A wall would prevent people beeing able so see the drone directly and covered with a pretty metal mesh, it would even reduce the wifi / radio waves passing through, what could shorten the control range. For most non terroristic acts, that might help.

  21. As already mentioned above, radars for detecting small UAVs already exist and are actively being deployed at events in the US and overseas. I’ve personally seen units deployed here in Asia. The spinning propellers modulate the radar signal, making a very nice way to differentiate between a drone and a bird.

    Also as mentioned above, if you’re flying a preprogrammed path for nefarious actions, a telem or control link will most likely not be used so sniffing for RF is pointless.

    1. that should only work with carbon fiber props as wood and plastic (being dielectric) are transparent to radar so its impossible for them to modulate radar signal in any meaningful way. you have to remember these people are selling the illusion of safety so they are really over hyping what they can do.

  22. Drones as every high-powered electric device has a peculiar Magnetic Anomalies Signature, it could be possible to build a network of AI powered M.A. detectors trained to discriminate these drones from other M.A. and then triangulate its position and signal its presence warning, the practical issue to this system is to train the MAD’s AI with thousand drones on the market (or not the drones, its motors are the MA sources), also malicious perpetrators could magnetically armor its drones and fool the system but at least non-malicious perpretators could be easyly spotted and shutdown by usual means.

  23. The problem is that it is certainly possible that a drone could be controlled by 4G or GPS waypoints. These things annot be blocked without causing problems on the ground. Detection systems are pretty worthless without a response system. AIrports are not going to shoot these things down because of safety issues in an urban area. I like an earlier poster’s suggestion of a “mechanical eagle” solution. The only effective way to deal with these is a hunter-killer autonomous drone. This tech has not been developed yet, but probably will!

  24. The two characteristics most easily identified, on a commercial/hobby drone, are the sound they create and the radio frequencies they use.

    It’s not hard to come up with sensors to detect both

  25. The 2.4 GHz is the common RF used by Quadcopters for connecting the ground transmitter to the drone.
    So we can use Anti-Drone Jammers by implementing Anti-UAV(unnamed aerial vehicle) Defense System (AUDS) It scans the skies for drones and jams their control signals using its own high-powered radio signal.

  26. With another Drone possible incident today at Heathrow, lets try and summarise possible solutions

    1. Sound Detection – Pros: Passive Cons: May not work against airport ambient noise, weather dependent, difficult to cover something the size of an airport, high false positives
    2. Optical Detection – Pros: Passive Cons: Similar to sound, plus visual detection is very directional, and unlikely to work at night
    3. IR Optical Detection : Similar to optical, but perhaps with better target discrimination. However may not work well in daylight or poor weather
    4. Lidar: Pros: Reasonable Range, will not effect existing systems Cons: Range, false detection rate (i.e Birds), weather dependent (Mist/Fog)
    5. Signal Analysis Pros: Passive cons: Would not stop GPS routed drones, false detection due to noisy EM space, poor direction and ranging information
    6. Radar Pros: Existing solution, good range, possible to discriminate against birds due to stregnth of returns and doppler discrimination Cons: May affect existing radar systems, may not detect low flying drones due to ground clutter
    7. Force detection apparatus onto drone such as radar reflectors or IFF Pros: Effective Cons: Easy to remove and IFF hardware may be short range and bulky. In effect, same problem as with present geo-fencing

    Maybe the best solution would be a combination. Radar, and signal analysis slaved to optical detection. Big concern would be intererfence of the radar system with things like ATC or communications. Maybe number of low power radar transmitters around the airport perimeter, slave to a large optical system. Maybe a large number of radio masts, emitting low level EM, which if disturbed by a metal object flying through it, is detected via the disturbance in the the transmission paths

    Once we have detected it the next question what we do not

    1. signal Jamming – Pros Easy to use Cons: Short Range due to power requirements, possible interference with domestic and commercial systems, won’t work on automated systems
    2. Shooting it down with a gun – Pros Present technology Cons: Requires expertise, danger of collateral damage, very hard to do with any confidence, especially at medium to long range
    3. Missile system with net – Pros Less danger of collateral damage Cons: Expensive, poor success rate, short range
    4: Force down via helicopter – Pros good Range,little danger of collateral damage Cons: Slow response time, Drones are small targets so hard to spot
    5.Eagles Pros:- good Target Discrimination Cons: Birds of prey need preparation to be made ready to fly, difficult to train, large drone may hurt a bird

    To be honest there is no good way to bring a drone down. Maybe the most likely solution would be to set a thief to catch a thief, so a hunter killer drone, initially slaved to the detection system and with short range optical/radar sensors to lock onto and disable the drone either via collision or some sort of net

    1. That’s a good synopsis of the situation – There’s no good solution with adequate precision, accuracy and collateral damage …….. Perfect challenge for the HackadayPrize2019 !!!!!

Leave a Reply to DustinCancel reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.