Hacker News new | ask | show | jobs
by _alxk 1727 days ago
They apparently used some ML to make predictions about the position of the driver and the future position of the vehicle to account for the lag when they pulled the remote trigger from hundreds of miles away.
3 comments

Sounds like something you’d use a Kalman filter for.

Except maybe because a Kalman filter expects Gaussian inputs, something fancier is used to deal with nonlinear inputs?

https://en.m.wikipedia.org/wiki/Kalman_filter

That's pretty trivial 'AI'. The headline mentioning AI is just for clickbait, in reality it was just a remotely-operated drone gun with a fire control system.
Most 'AI' is trivial. If spam detection had been invented today they'd call it "Spam AI".
You're not wrong. I'm gonna add AI to our pitch deck for our A probably will be able to raise an additional 40%.
Right. As well, the gun mounted in the truck needs to account for real world wind, vehicle recoil/ follow through