Accelerometer May Help Make You A Sharpshooter

[Chris Suprock] is interested in using technology to improve your accuracy with a firearm. To that end, he’s using an Accelerometer mounted to a gun to gather feedback about each shot.

The hardware setup is pretty simple. We don’t have specific details, but it looks like he’s using a QFN accelerometer chip like you would find in a cellphone. The milled aluminum mounting bracket that holds the board has ‘USB’ printed on it, although the connector is something we don’t really recognize.

In the video after the break [Chris] demonstrates the feedback he can get when the device is mounted on the stock of a Ruger Mini-14. The graph of the data makes it obvious when the trigger was pulled. The most useful part may be the period leading up to that event, as it shows any unnecessary movement prior to the shot. If you’re into sport shooting, this may be one more tool that will help give you the edge on your competitors.

36 thoughts on “Accelerometer May Help Make You A Sharpshooter

    1. Problem is most long range rifles only have a top rail, and that is usually completely taken up by the scope. A scope mount in the form of rings would be more appropriate.

      1. That would be easy to fix. Use a style similar to a Nightforce Angle Degree Indicator that mounts on the rail underneath the scope, and hangs off the side, or mounts to the scopes body. That is also ignoring that you could hollow out the stock, part of the grip, etc.

    1. As far as I know, that has already happened. I remember a very old engadget(sic!) article where someone build a image stabilizer with an accelerometer and some lens-shifting… But I also know about an Article where they tested the effect of the shutter release to imagequality. I think it was dpreview who did this test, as they came to the fact, that the shutter release itself gives the image _always_ a bit of motion and therefore is not always really sharp. But… its was only visible in extrem comparison by subpixels and when you know what you were looking for.

    2. Firearms are fine if properly taken care of with regular maintenance and cleaning, not to mention properly stored, locked and obviously used, it’s the idiots using them that I hate… A gun is a tool, and much like other tools (such as a knife, a chainsaw, a lathe, hell, even a car) they can be deadly if misused… The problem with firearms is that people think they’re “cool” and treat them as toys to play around with, and that they NEED one even if they never plan on using it for its intended purpose. Don’t blame the tool for the user’s idiocy.

      1. EXACTLY! I hate it when people say things like “Guns kill people”. You could say the same thing about pretty much everything. Guns don’t kill people, people kill people.

      2. Oh yeah I forgot, one more thing. Guns are cool. And everyone in the USA has the right to own one. I believe that guns are just great to have in the household (if properly secured in a safe or other means of protection) for safety and self defense reasons. I believe that people should absolutely not treat them like toys but rather like tools, responsibly.

  1. Nooooooooooooo! I’ve been working on interpreting data from an accelerometer attached to a rifle also, in order to analyze the shooter’s breathing and trigger squeeze mechanics, as well as the rifle’s recoil characteristics. Late to the party again, I guess. Good job, Chris.

    1. Your project is only obsolete if you think you wouldn’t leard anything anymore by continuing it and/or your project would not get to the level of his project. What could you add to the Project that he didnt thought of? In other words: CONTINUE!

      1. Indeed. What would be awesome is heuristic feedback during the entire shooting process. Just some simple lights that tell you that something isn’t right and you may miss.

    1. Oh goodness, for that price you could make about a dozen of those. The implementation is still interesting though. I was actually wondering “Why didn’t he make this wireless?” but I guess that answers it.

    2. As the engineer that designed this product, I feel obligated to respond to this comment. The accelerometer is a harsh environment sensor designed to measure part quality and tool wear in high volume metal cutting applications.

      The hack is Chris using this accelerometer to do something neat in an entirely different application from the one for which it was designed.

      Jeff Nichols,
      Suprock Technologies

    3. As someone who designed a similar gadget at my day-job, I recognize those accelerometer specs ;-) That is an ADXL345 onboard; they run about $3 in high quantity. (Of course, there is a lot more to the product than the accelerometer, so $hundreds$ price tag for the finished gadget + support + hw/sw development costs is not so unreal at small volumes.)

      One thing I’m not clear on is how (or if) they trim the timing of the output data (ADXL345 generates its own sample clock; 3200Hz is the *nominal* value, which varies a not-insignificant amount part to part and over temperature). More of a concern for frequency-domain analysis, but it’s very cheap (or free) to do in software and turns the ‘345 into a pretty decent part.

  2. I am now dreaming about what can happen if you combine this with:

    1. LEDs
    2. Convex lens with >20mm focal length
    3. Half-silvered glass or beam splitter

    (hint: aircraft HUDs work like this)

    Indeed, I’m imagining feedback being displayed directly into the scope picture, in realtime. Place the above contraption in front of the optics…

  3. It’s actually not a QFN package. It’s an LGA package (the pads are completely underneath; think BGA without the balls). QFNs are significantly easier to solder. Even the BGA on that particular board didn’t give us nearly so much trouble during our early prototyping.

    Jeff Nichols,
    Suprock Technologies

  4. Rifle accuracy is pretty trivial compared to handgun accuracy. This would be much more useful (though still of dubious usefulness, I think) mounted on a handgun to collect data about flinch and trigger jerk.

  5. @loans: rifle shooting is trivial compared to pistol? Really. I instruct and shoot both competitively, and work to achieve 1/2 MOA, so ANYTHING that can inform me of pre-shot movement IS going to help. Even the movement after discharge is important and can help me develop the ideal position. Shooting is all about position and repeatability. The difference in scoring is a fraction of a minute of angle, and we shoot out to 600 yards. Everything matters. (OTOH, shooting deer at 50 yards is certainly not all that challenging, from a teargeting perspective.)

  6. The peak rifle acceleration saturates the signal as can be seen from the clip of the waveform. The mounting plates from the very first clip are labeled +/-16g and +/-24g. The weapon’s peak recoil could be 100g’s or more. The amplitude of the signal (the g-force “readout”) is unreliable as it is likely being resonanted from the shock of the weapon beyond the sensor’s resonant frequency, the system is not fast enough to capture the event (3,200 samples per second) when a high-powered rifle’s rearward acceleration is half-over in the time it takes this system to capture two data points. The graph also needs to be “zeroed” as the green channel is base-lined at 1g.
    This system demonstrated here is only good enough to make assumptions rather than be used to document results. Data Validation, that’s key. You could use this to monitor the low frequency “anticipation” of the rifleman but not the rifle dynamics and nothing substantial after the event.

  7. Great idea with helping ‘fight’ the pre-shot anticipation with pistols! The hard ones are those long DAO pulls, especially after putting away the 1911 for the day. After getting “rid” of it with the 1911, it comes right back with the DAO (to me!) every time. :(

Leave a 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.