Seeking suggestions and advice on MMW doppler radar to triangulate ballistic projectiles for target scoring.

Thread Starter

Moonshot

Joined Mar 17, 2021
16
I am an avid marksman in both archery and firearms and have been considering for sometime on creating a target scoring device that can be used for both hobbies. Specifically, I am interested in a technology that can detect the ballistics of projectiles traveling at both subsonic and supersonic speeds. The primary focus of the device is to triangulate, with high accuracy, the impact location of the projectiles after hitting a target. This device would be located safely below the target with the radar cone ideally aligned with the x and y axes of surface of the target. The device will need to detect the projectile and triangulate its virtual impact position within a minimum 1 meter square target area and up to 2.5 meters square. It would also need to accurately detect impacts as near as 24 centimeters to itself. I would also like to have the accuracy to within 3mm or better and finally as an option; the means to calculate the speed of the projectiles at the target.

I have experience with scoring systems that rely on ultrasonic microphones to detect bullet impacts and are accurate to within 3mm but these only work with supersonic projectiles as they detect the supersonic cone for triangulation. Another draw back is that the microphones are exposed line of sight and make them vulnerable to destruction by the projectiles.

I am assuming that doppler radar would be a superior means of triangulation and speed detection, so my thought is this is the way to go. Are there any off the shelf sensors that may be used for this purpose? Perhaps something DIY friendly that can be used in conjunction with an Arduino type wireless board so the data can be accessed and translated by a software algorithm for visual representation on a screen.
 

Yaakov

Joined Jan 27, 2019
2,342
I would think, with the current state of deep learning AI, and optical approach would be easier. Just a thought.
 

Papabravo

Joined Feb 24, 2006
15,453
I'm thinking the time scales and the low radar cross section of a speeding bullet would make this a near impossible task. It is one thing to track an airframe 100 miles away and quite another to track something that small over such a short distance. Maybe somebody with more radar experience than I have can weigh in. My rag wing Citabria and I were completely invisible to Cleveland Center from 20 miles out until I turned my propeller arc in their direction. No, I was not inside the TCA boundaries, I just wanted them to be aware of my presence.
 

Thread Starter

Moonshot

Joined Mar 17, 2021
16
I'm thinking the time scales and the low radar cross section of a speeding bullet would make this a near impossible task. It is one thing to track an airframe 100 miles away and quite another to track something that small over such a short distance. Maybe somebody with more radar experience than I have can weigh in. My rag wing Citabria and I were completely invisible to Cleveland Center from 20 miles out until I turned my propeller arc in their direction. No, I was not inside the TCA boundaries, I just wanted them to be aware of my presence.
The type of doppler radar that I believe applies to my situation would be millimeter wave (microwave) in the 24ghz to 77ghz frequency range which is largely used in automobile and robotics. LabRadar is a product that actually tracks the speed of bullets to about 20 meters with a 24ghz transciever. So I'm fairly confident this is where I should be looking for a solution.
 

Papabravo

Joined Feb 24, 2006
15,453
The type of doppler radar that I believe applies to my situation would be millimeter wave (microwave) in the 24ghz to 77ghz frequency range which is largely used in automobile and robotics. LabRadar is a product that actually tracks the speed of bullets to about 20 meters with a 24ghz transciever. So I'm fairly confident this is where I should be looking for a solution.
The expense of working in that frequency regime will produce quite a case of sticker shock.
 

Yaakov

Joined Jan 27, 2019
2,342
I am really not trying to be negative but I can see so many problems with this approach. First, you will have almost no time to get the data. Millimeter waves or not, if the device is one foot from the target, let's say, and the projectile is moving at 1000 fps, you will have 1/1000s worth of data. You will have to collect it very fast which means some very expensive ADCs.

Additionally, there is the issue of calibration. The target's exact location relative to the device will need to be known, which means some sort of registration and only bespoke targets.

The optical thing seems better and better to me. It would be self calibrating and the data is static. Alternatively, acoustic or pressure sensors on the backboard would allow for triangulation and while calibration would still be an issue, the data would be much easier to collect.

I could be wrong, but I think you might want to rethink the technology before it becomes the problem instead of the solution.

Again, this isn't meant to be negative, it's just my (possibly jaundiced) professional eye making a first approximation.
 

Thread Starter

Moonshot

Joined Mar 17, 2021
16
I am really not trying to be negative but I can see so many problems with this approach. First, you will have almost no time to get the data. Millimeter waves or not, if the device is one foot from the target, let's say, and the projectile is moving at 1000 fps, you will have 1/1000s worth of data. You will have to collect it very fast which means some very expensive ADCs.

Additionally, there is the issue of calibration. The target's exact location relative to the device will need to be known, which means some sort of registration and only bespoke targets.

The optical thing seems better and better to me. It would be self calibrating and the data is static. Alternatively, acoustic or pressure sensors on the backboard would allow for triangulation and while calibration would still be an issue, the data would be much easier to collect.

I could be wrong, but I think you might want to rethink the technology before it becomes the problem instead of the solution.

Again, this isn't meant to be negative, it's just my (possibly jaundiced) professional eye making a first approximation.
I really appreciate your input. You have answered several questions that I had about the technology but didn't ask directly. i,e, time, cost and calibration. I wouldn't really be interested in the optical route, if by that you mean with incorporating cameras and microphones to triangulate for scoring automation. I may just purchase a super sonic acoustic system by Shot Marker ($800) for rifles and figure out how to augment standard microphones for the sub sonic triangulation.
 

Yaakov

Joined Jan 27, 2019
2,342
I really appreciate your input. You have answered several questions that I had about the technology but didn't ask directly. i,e, time, cost and calibration. I wouldn't really be interested in the optical route, if by that you mean with incorporating cameras and microphones to triangulate for scoring automation. I may just purchase a super sonic acoustic system by Shot Marker ($800) for rifles and figure out how to augment standard microphones for the sub sonic triangulation.
Well, I meant a camera pointed at the target (it can be close and below and using deep learning to spot the shots but literally seeing them on the target. That way, it could be trained for any number of targets. Just an idea.

If the microphones were in contact with the backing board, you could use time of flight triangulation based on the round hitting the board. Alternatively, you could use force sensors and have them mount the frame for the backing board and do the same thing.

However, if there is an existing solution, you may well be able to use it as the basis for subsonic scoring as well.
 

Reloadron

Joined Jan 15, 2015
5,875
Have you looked at Lab RADAR? While it won't give you target point data it will use doppler RADAR to track a projectile accurately out to 100 yards downrange. While I am not using it I remember it was a long time in development and over the past few years it gets good reviews. Being of dinosaur nature I still use a Oehler 35P for velocity and a spotting scope. :)

Ron
 

wayneh

Joined Sep 9, 2010
17,024
Complete outsider here. Maybe out of the box thoughts are useful, probably not.

It seems to me that determining where a round hit and how fast it came in are vastly different tasks. You have plenty of time after the impact to answer “where” with sub-mm precision.

Speed at the target is entirely different. As a fan of “Smarter Every Day”, I’d probably look at high speed photography as a possible answer. (It would also show where a projectile hit as a bonus.) There’s no substitute for actually seeing an event with your own eyes.

But I get it. You want the equivalent of the mph meter at the baseball pitching booth at the county fair. Immediate feedback of on target accuracy and pitching speed. It’d be easier if the round was always precisely on target!
 

Danko

Joined Nov 22, 2017
1,060
The primary focus of the device is to triangulate, with high accuracy, the impact location of the projectiles after hitting a target.
Here is concept of metal target 1200x1200mm, PDF.
And C++ program, ZIP.
Sorry, it is in Russian.
ADDED
file target_AAC_exe.zip
 

Attachments

Last edited:

andrewmm

Joined Feb 25, 2011
1,128
Re radar,
you are either looking at a pulsed system, where you measure the time of flight / return
or a CW one, where you transmit continuous and measure the doplar shift of the received signal

yes there are mixtures of the two,

without encoding, the doplar can not detect range to target, only its there, and its speed,
without a narrow beam, you can not detect direction accurately,

a narrow beam implies an antenna with gain,

The pulsed radar,
the time between pulses puts a limit on the distance you can see,
and the length of the pulse limits how accurately you can range bin the target,

just about everything you might want to know abut radar

https://www.radartutorial.eu/html/sm03.en.html
 

Papabravo

Joined Feb 24, 2006
15,453
Re radar,
you are either looking at a pulsed system, where you measure the time of flight / return
or a CW one, where you transmit continuous and measure the doplar shift of the received signal

yes there are mixtures of the two,

without encoding, the doplar can not detect range to target, only its there, and its speed,
without a narrow beam, you can not detect direction accurately,

a narrow beam implies an antenna with gain,

The pulsed radar,
the time between pulses puts a limit on the distance you can see,
and the length of the pulse limits how accurately you can range bin the target,

just about everything you might want to know abut radar

https://www.radartutorial.eu/html/sm03.en.html
There are immense difficulties in designing and operating RF equipment in that frequency range. If you check prices on RF equipment for that frequency range you will be amazed at what you have to spend. It is small wonder that only the military and a few dedicated amateur enthusiasts dabble in this realm. One expense item is the 10x rifle scope to align their antennas with the target of their communications attempts.
 

Thread Starter

Moonshot

Joined Mar 17, 2021
16
Complete outsider here. Maybe out of the box thoughts are useful, probably not.

It seems to me that determining where a round hit and how fast it came in are vastly different tasks. You have plenty of time after the impact to answer “where” with sub-mm precision.

Speed at the target is entirely different. As a fan of “Smarter Every Day”, I’d probably look at high speed photography as a possible answer. (It would also show where a projectile hit as a bonus.) There’s no substitute for actually seeing an event with your own eyes.

But I get it. You want the equivalent of the mph meter at the baseball pitching booth at the county fair. Immediate feedback of on target accuracy and pitching speed. It’d be easier if the round was always precisely on target!
I have almost no desire for speed detection. Triangulation is really the main goal; I can track speed with any chronograph. You seem to imply that the task would be much more feasible taking the speed part out of the equation... Anyone who can add to the concept please feel free to reply.
 

Yaakov

Joined Jan 27, 2019
2,342
I have almost no desire for speed detection. Triangulation is really the main goal; I can track speed with any chronograph. You seem to imply that the task would be much more feasible taking the speed part out of the equation... Anyone who can add to the concept please feel free to reply.
I don't see any easing of the fundamental problems you have to solve to make it work. If you could make it work, it would come at an enormous cost. There will be no economy of scale and the hardware and software aspects would be extremely fiddly.

I don't think it is a practical idea and as far as I can tell it doesn't offer any advantages over other ways of doing it.
 

Thread Starter

Moonshot

Joined Mar 17, 2021
16
Conclusion
detecting and tracking a bullet is not an easy task

in military systems, we can do it for shells and mortar rounds,
but bullets are limited to sound detection as to where the shot came from

https://en.wikipedia.org/wiki/Gunfire_locator#:~:text=A gunfire locator or gunshot,a combination of such sensors.
I would like to clarify my intention. The goal is not to track or to measure speed. The goal is to triangulate a projectiles impact on the surface of a target.
Conclusion
detecting and tracking a bullet is not an easy task

in military systems, we can do it for shells and mortar rounds,
but bullets are limited to sound detection as to where the shot came from

https://en.wikipedia.org/wiki/Gunfire_locator#:~:text=A gunfire locator or gunshot,a combination of such sensors.
I am an avid marksman in both archery and firearms and have been considering for sometime on creating a target scoring device that can be used for both hobbies. Specifically, I am interested in a technology that can detect the ballistics of projectiles traveling at both subsonic and supersonic speeds. The primary focus of the device is to triangulate, with high accuracy, the impact location of the projectiles after hitting a target. This device would be located safely below the target with the radar cone ideally aligned with the x and y axes of surface of the target. The device will need to detect the projectile and triangulate its virtual impact position within a minimum 1 meter square target area and up to 2.5 meters square. It would also need to accurately detect impacts as near as 24 centimeters to itself. I would also like to have the accuracy to within 3mm or better and finally as an option; the means to calculate the speed of the projectiles at the target.

I have experience with scoring systems that rely on ultrasonic microphones to detect bullet impacts and are accurate to within 3mm but these only work with supersonic projectiles as they detect the supersonic cone for triangulation. Another draw back is that the microphones are exposed line of sight and make them vulnerable to destruction by the projectiles.

I am assuming that doppler radar would be a superior means of triangulation and speed detection, so my thought is this is the way to go. Are there any off the shelf sensors that may be used for this purpose? Perhaps something DIY friendly that can be used in conjunction with an Arduino type wireless board so the data can be accessed and translated by a software algorithm for visual representation on a screen.
[/QUO
I am an avid marksman in both archery and firearms and have been considering for sometime on creating a target scoring device that can be used for both hobbies. Specifically, I am interested in a technology that can detect the ballistics of projectiles traveling at both subsonic and supersonic speeds. The primary focus of the device is to triangulate, with high accuracy, the impact location of the projectiles after hitting a target. This device would be located safely below the target with the radar cone ideally aligned with the x and y axes of surface of the target. The device will need to detect the projectile and triangulate its virtual impact position within a minimum 1 meter square target area and up to 2.5 meters square. It would also need to accurately detect impacts as near as 24 centimeters to itself. I would also like to have the accuracy to within 3mm or better and finally as an option; the means to calculate the speed of the projectiles at the target.

I have experience with scoring systems that rely on ultrasonic microphones to detect bullet impacts and are accurate to within 3mm but these only work with supersonic projectiles as they detect the supersonic cone for triangulation. Another draw back is that the microphones are exposed line of sight and make them vulnerable to destruction by the projectiles.

I am assuming that doppler radar would be a superior means of triangulation and speed detection, so my thought is this is the way to go. Are there any off the shelf sensors that may be used for this purpose? Perhaps something DIY friendly that can be used in conjunction with an Arduino type wireless board so the data can be accessed and translated by a software algorithm for visual representation on a screen.
I don't have the ability to edit the post so I would like to clarify the concept. The goal is not to track a projectile. Speed detection is also not important but as i mentioned in the post that it was optional if taking speed detection out of the equation makes it easier then ignore it. The main goal is to triangulate a projectiles position at impact on a target with a radar device in close proximity and below the line of site of the intended target. The radar would detect the passing projectile as it breaks the planar between the transceiver and the target face. The data would then be sent wirelessly to a server that would perform triangulation with an algorithm and display the impacts on a visual device.
Here is concept of metal target 1200x1200mm, PDF.
And C++ program, ZIP.
Sorry, it is in Russian.
Thanks! I'll have it translated.
 

Thread Starter

Moonshot

Joined Mar 17, 2021
16
I don't see any easing of the fundamental problems you have to solve to make it work. If you could make it work, it would come at an enormous cost. There will be no economy of scale and the hardware and software aspects would be extremely fiddly.

I don't think it is a practical idea and as far as I can tell it doesn't offer any advantages over other ways of doing it.
@Yaakov, Bringing me back down to earth! LOL. However, after reading Wayneh's response I can't get it out of my mind. A projectile breaks the planar of the radar traveling at 2000 fps while the radar is transmitting a continuous wave 24ghz beam. The beam is traveling near speed of light, which reflects back to the multiple receiving antenna. Regardless of the exposure time and given the relative slow speed of the projectile; significant pulses are generated from the reflections and data created with time of event and distance to projectile from two or more antennas. It just seems intuitive that there is enough data to work with and the hardest part is the math, the algorithm, the interface. Of course R&D related to this would certainly be a pricey obstacle but might be mitigated with open source development. I'm probably being naïve but I hade to get it out of my system.
 
Top