06-01-2024, 10:31 AM
https://www.yahoo.com/news/ukraines-spec...48694.html
Hallelujah! About time! Now we just need to make sure they have enough of these items. It would also be "nice" to utilize this technology on larger drones that can carry a larger warhead, fly longer ranges, and search larger areas well behind Russian lines. This would not be all that expensive for NATO to do either. and it uses components we make.
To be clear, this type of technology isn't all that "new" in itself. The AARGM utilizes this type of technology to search for enemy radars that have shut down. The infrared CCD imaging system and the W band radar both contribute to the solution. The Storm Shadow also has image recognition mode in its terminal guidance. What is "innovative" is creating a system to do this that fits on a small drone. The SWAP requirements are a big challenge, and don't be too surprised if it has to go through iterative improvements to make it work "well". Ukraine is probably using "alpha" versions right now. There is really no adequate way to test these weapons on a test range. Field testing is essential to work out design flaws in the recognition system. I hope these drones have a "backlink" so Ukraine can evaluate what they actually do in "real life". Cramming all that into the drone may be impossible however.
IMO an anti-radiation drone is a much better solution to the destruction of GPS jammers. I am continually amazed that Ukraine hasn't developed that capability. It is MUCH easier than developing a robust (possibly multi-spectrum) image recognition approach. It is also longer range. Perhaps they encountered some problem I am not aware of, but it still seems strange.
There are a couple of problems with using "small" drones to kill GPS jammers. One is that the jammer can be quite a distance away from the front line. The drone has to be able to fly that far. Second, searching for a jammer may require searching a wide area. If no geolocation or (at a minimum) line of bearing information is available it seems very improbable the "multicopter" drones have the endurance to reliably "find" the jammers. If a line of bearing (or geo) is available the drone can fly a dead-reckoning course that will get it in the immediate neighborhood. The probability of finding a jammer by image recognition is thereby greatly improved. The DR plot can be improved by using a compass and accelerometer (https://www.adafruit.com/product/1120) to compensate for "windage". It isn't good enough to qualify as an "INS" but it doesn't need to be.
Decoys are a serious problem for image recognition systems. A bunch of plywood GPS jammer antennas can make it very difficult to find the "real" ones. That is why "military grade" image recognition systems are almost always multi-spectrum. Building something that will "fool" the software in both/many spectra isn't at all easy. It does however increase the SWAP requirements and the $ requirements. There is no free lunch.
Hallelujah! About time! Now we just need to make sure they have enough of these items. It would also be "nice" to utilize this technology on larger drones that can carry a larger warhead, fly longer ranges, and search larger areas well behind Russian lines. This would not be all that expensive for NATO to do either. and it uses components we make.
To be clear, this type of technology isn't all that "new" in itself. The AARGM utilizes this type of technology to search for enemy radars that have shut down. The infrared CCD imaging system and the W band radar both contribute to the solution. The Storm Shadow also has image recognition mode in its terminal guidance. What is "innovative" is creating a system to do this that fits on a small drone. The SWAP requirements are a big challenge, and don't be too surprised if it has to go through iterative improvements to make it work "well". Ukraine is probably using "alpha" versions right now. There is really no adequate way to test these weapons on a test range. Field testing is essential to work out design flaws in the recognition system. I hope these drones have a "backlink" so Ukraine can evaluate what they actually do in "real life". Cramming all that into the drone may be impossible however.
IMO an anti-radiation drone is a much better solution to the destruction of GPS jammers. I am continually amazed that Ukraine hasn't developed that capability. It is MUCH easier than developing a robust (possibly multi-spectrum) image recognition approach. It is also longer range. Perhaps they encountered some problem I am not aware of, but it still seems strange.
There are a couple of problems with using "small" drones to kill GPS jammers. One is that the jammer can be quite a distance away from the front line. The drone has to be able to fly that far. Second, searching for a jammer may require searching a wide area. If no geolocation or (at a minimum) line of bearing information is available it seems very improbable the "multicopter" drones have the endurance to reliably "find" the jammers. If a line of bearing (or geo) is available the drone can fly a dead-reckoning course that will get it in the immediate neighborhood. The probability of finding a jammer by image recognition is thereby greatly improved. The DR plot can be improved by using a compass and accelerometer (https://www.adafruit.com/product/1120) to compensate for "windage". It isn't good enough to qualify as an "INS" but it doesn't need to be.
Decoys are a serious problem for image recognition systems. A bunch of plywood GPS jammer antennas can make it very difficult to find the "real" ones. That is why "military grade" image recognition systems are almost always multi-spectrum. Building something that will "fool" the software in both/many spectra isn't at all easy. It does however increase the SWAP requirements and the $ requirements. There is no free lunch.