I wonder how long until DIY weaponized drones are available to common criminals.
It doesn't seem like we are prepared to deal with something like this.
It doesn't seem like we are prepared to deal with something like this.
Common criminals are using commodity drones for reconnaissance [1].
[1] https://www.techradar.com/news/criminal-intent-fbi-details-h...
thinking further, even just the manner, or tactics used when flying could be weapon like, such as chaff drones
It’s trivial, cheap, and simple to do massive harm to massive numbers of people.
Fortunately, humans don’t usually do this unless part of the government/military.
In a country of 300 million, the only non-state mass death events are a handful of terrorist attacks and deranged teenagers determined to shoot up their schools. The annual deaths caused by these non-state actors is, on average, less than the annual fatalities on a given stretch of a run-of-the-mill interstate.
In 2020, 446 people died in mass shootings in the US. [2]
1. https://komonews.com/news/local/seattle-traffic-deaths-remai...
2. https://en.wikipedia.org/wiki/List_of_mass_shootings_in_the_...
In any event >40k people died from traffic fatalities in the United States in 2020[0], so a few orders of magnitude larger than mass shooting deaths,
[0] https://www.npr.org/2021/03/05/974006735/tragic-driving-was-...
I wonder if standard BBs or ball bearings could achieve sufficient terminal velocity to pierce skin - would mitigate accuracy requirements if you could dump them by the dozens or hundreds.
If you are using a wing, then you just program your robot to come in high, cut it's engine and continue on a glide trajectory to the target. The flap servos make nearly no noise.
If you've got a quad, come in fast and basically huck the vehicle at the target.
Ardupilot + ROS make this all relatively straightforward to implement and provide an easy way to interface with necessary sub-systems.
To do this with any reliability you need custom software and probably hardware. It'll look more like a smart munition. But it'll also be more difficult than strapping a knife to a RC plane.
Second, who said the quad was tumbling? Significantly cutting engine speed and removing significant lift forces on a quad in order to kill the buzz doesn't mean it will start wildly gyrating.
Third, Ardupilot will get you within a few feet of a specified point with a good enough vector and speed. Luckily you can use a subroutine (secondary control program) to handle the final approach from the 'point' to the target.
A quad will tumble with the motors off, the whole thing is a significant lift force. It was a big problem in early controller software where throttle dead zone would turn off the props. A problem I personally experienced. Maintaining any semblance of stability requires active adjustment.
And no, Ardupilot will not get anything going at high speed to within a few feet in 3d space.