5 ms·
Seems like small drones are the best weapon against big drones. Kind of curious that they seem to assume that the drone needs an operator. If I were defending a
by Aeolun 4y ago
Seems like small drones are the best weapon against big drones. Kind of curious that they seem to assume that the drone needs an operator. If I were defending against a drone swarm I’d want my drones to be piloted by AI.
- adrianN 4y agoPerfect micro matters a lot when you zergrush your enemies: https://www.youtube.com/watch?v=IKVFZ28ybQs https://www.youtube.com/watch?v=IKVFZ28ybQs
- mgdlbp 4y agoAlphaStar's blink stalker micro, https://www.youtube.com/watch?v=cUTMhmVh1qs&t=5888s https://www.youtube.com/watch?v=cUTMhmVh1qs&t=5888s with capped actions-per-minute and reaction time: https://news.ycombinator.com/item?id=18992698 https://news.ycombinator.com/item?id=18992698
- MaxikCZ 4y agoThis is what I dont like about Starcraft. The strategy is, in many instances, second to microing the units. The fact that 1hp unit have the same effectivenes as a full hp one is just weird.
- picsao 4y agoCarrier drones are the ideal, swallowing cost of transport while not sacrificing amount and speed
- 323 4y agoSmall drones have limited range, so you need to deploy them everywhere. The front line in Ukraine is 1000 km. Maybe you can have a big "carrier" drone transport them to where they are required, like in StarCraft. Or maybe you can make the small drones recharge directly from the grid, by landing on top of power lines and deploy the necessary connections.
- jonathankoren 4y agoSmall attritable drone swarms deployed by strike aircraft. https://youtu.be/ndFKUKHfuM0 https://youtu.be/ndFKUKHfuM0 https://en.wikipedia.org/wiki/Perdix_(drone) https://en.wikipedia.org/wiki/Perdix_(drone)
- usrusr 4y agoIf you want a defensive screen: minimalistic gliders, deployed to high altitudes, that dive down onto a target when needed. They'd simply use altitude as energy storage, and when that's nearly used up they take the rest to make it to some designated landing field, ready to get collected for the next launch. Multiple "shifts" at different points in the altitude cycle at any given time.
- nradov 4y agoAnd what kind of sensors do those gliders have, and how much power do those sensors draw?
- usrusr 4y agoSensors fly separately from the consumables, hitching the same ride up. The consumable ones only need their own for final approach which might even be assisted by "painting". A bigger problem would be sensor overload from all the friendlies, you'd probably want those loitering to follow a course determined by a PRNG predictable to everyone who knows that day's key. But that's just very basic crypto put to novel use.
- nradov 4y agoThat depends on the flight regime. If the big drones are at high altitude and/or moving fast then a small drone will never have the performance or sensors to achieve an intercept. So we need fast missiles or directed energy weapons fired from large, expensive platforms (which could be manned or unmanned, operating on land, sea, or air). AI piloting technology remains very primitive. We don't have anything yet that can cope well with a dynamic air combat environment involving multiple unpredictable threats, EW, countermeasures, and complex ROE. That's just total science fiction today so we can expect to have humans in the loop at some level for many years to come.
- Aeolun 4y agoI think you can generally build a smaller drone than the attack drone, regardless of environment, if your goal is simply to trade them 1 for 1 (e.g. suicide drones).
- nradov 4y agoWhat do you mean by drones exactly? If you're referring to small multi-rotor UAVs, those are so slow as to be almost completely useless as interceptors. Need something faster, with rocket or turbine propulsion. Probability of kill is never going to be 1.0, so the defender will usually fire multiple interceptors per target.
- ajuc 4y agoPiloting isn't the main bottleneck. It's easier to automate the piloting than the maintenance, and unless we're talking one-use drones (so, guided missiles really) - there's a lot of maintenance. Somebody needs to transport the drones to the front, program the goals, recharge the batteries (which takes hours often), rearm them after the mission, download the gathered data, change the encryption keys, etc. And if you manage to automate some of that - automation make your drone swarm vulnerable to disruption. Enemies can wait near automatic refueling/recharging station and capture the drones one-by-one for example. They can put a gps tracker on the automatic drone transport system and shell the enemy base once it gets there. Or they can put a grenade on a captured drone, release it, and detonate it when the drone is recharging destroying the whole swarm cheaply. Ultimately there's a trade-off between automation and resilience that should be familiar to web developers.
- antihero 4y agoWhat if you launch the drone swarms from a large plane circling out of danger, like an AWACS or gunship?
- squeaky-clean 4y agoAnything far enough away to not be targeted by a ground to air missile would also probably be too far away for a drone to make a round trip. A quick Google search says even SAMs from the 1950s can hit targets over 30km away.
- dinkledunk 4y agoPut your drones inside a missile and launch them at their target.
- mattlondon 4y agoI saw a slide somewhere recently - there are already plans for this as an artillery shell or missile. I think it was combined with ordinary ordnance- drone is ejected before main weapon payload towards end of "flight". Drone helps pick highest value target then is used after for damage assessment. Presumably then it self destructs or/or kamikazees
- bilsbie 4y agoActually it’s ordinary table yarn or fishing line. Fly it around the drones and trip up the props.