Hmmm, while this is very impressive demonstration on its own, one can't help but wonder about how much operational relevance it actually has. Sure the drone was "detected" 4km out. But only as a tiny blurry blob, and with the system staring directly at it. The website states the camera system's FOV is "8.8º-0.6º", so to get solid coverage of the entire horizon at that distance would require... a lot of these! I didn't see a price listed on the site, but I can't imagine that'd be cheap. Furthermore, even with the most advanced AI assistance, it must be incredibly difficult to identify a threat in all the noise of moving background, birds, bugs, turbulence, etc. So even having the drone in-frame wouldn't guarantee detection. The fact the video editor decided to use a circle to highlight the drone in these clips speaks to that. All in all, not really sure what this test proves in regards to dronetection. The Ventus could certainly be useful in conjunction with some sort of hybrid integrated sensor network, but that's not what was discussed above. What'd be fun and would prove a lot more is to zip-tie a paperweight to the bottom of a Mavic and see how close a pilot could get to directly over the observer before the drone was spotted. As the pilot and spotter got more experience and switched off, it'd get progressively more challenging for both as the pilot tried different tricks and the spotter got better at spotting. That'd get some actually relevant numbers.
How are these drones that are currently being spotted able to avoid thermal image detection?
You guys should add object dection software make class of different types of drones this will be powerful tool in our defense in sky
Hmmm, while this is very impressive demonstration on its own, one can't help but wonder about how much operational relevance it actually has.
Sure the drone was "detected" 4km out. But only as a tiny blurry blob, and with the system staring directly at it.
The website states the camera system's FOV is "8.8º-0.6º", so to get solid coverage of the entire horizon at that distance would require... a lot of these! I didn't see a price listed on the site, but I can't imagine that'd be cheap.
Furthermore, even with the most advanced AI assistance, it must be incredibly difficult to identify a threat in all the noise of moving background, birds, bugs, turbulence, etc. So even having the drone in-frame wouldn't guarantee detection. The fact the video editor decided to use a circle to highlight the drone in these clips speaks to that. All in all, not really sure what this test proves in regards to dronetection. The Ventus could certainly be useful in conjunction with some sort of hybrid integrated sensor network, but that's not what was discussed above.
What'd be fun and would prove a lot more is to zip-tie a paperweight to the bottom of a Mavic and see how close a pilot could get to directly over the observer before the drone was spotted. As the pilot and spotter got more experience and switched off, it'd get progressively more challenging for both as the pilot tried different tricks and the spotter got better at spotting. That'd get some actually relevant numbers.