I posted something on this in the trad heli forum, but there's been no feedback, and the user group forum seems to get more views, so I'll mention it here.
Recently, my robotics team has been playing with some pressure sensitive resistors, and it got me thinking:
One of the problems we tend to have is in stability while the heli is interacting with the ground. I was thinking that we could use some of these nifty pressure sensitive resistors to make touch sensors for the landing gear, then integrate them into the code to have a seperate routine run feedbeck while we're doing automated takeoffs and landings.
Pros: The sensors have enough dynamic range that we could potentially spring-load them and get a full range of voltages. Basically, using one sensors on each corner of the gear (or one on each foot for multicopters), we could have an additional feedback mechanism for ensuring that the heli takes off and lands with even pressure across the gear.
Cons:
1) It gives us one additional sensor network to have to calibrate.
2) It only works ideally for taking off and landing from flat, even surfaces. Even asphalt or gravel could have enough variation that the sensors could give garbled feedback if they're not particularly robust in their programming.
This is just something I was thinking about, as I have a few ideas for features that might be useful. If anyone else thinks this might be useful, then like the thread, and I'll get started on it after I finish helping Mick get HiL up and running.
Replies
I think this is a good idea. I have had discussions with a couple other folks with issues with landings and takeoffs with thier machines. One is is a fixed wing conversion aircraft and the other is a rotorcraft machine. Without information wether the aicraft is on the ground they tended to roll over sometimes at light ground contact or slow takeoffs and landings. A code patch/option would usefull.
I think it's a good idea. I can't guarantee that it will be taken up by the main code base but it'd be interesting to try out and prove it could work and see how it goes.
Certainly there are cases where we'd like to know if we're on the ground or not for example when landing! with sonar we know if we're within 20cm but that's it. Also we could use it to stop I terms from building up so we don't have wonky take-offs.
Another idea would be to use a flex sensor. Potentially if you mounted it straight down it would bend when you get close to the ground. It would need to be robust and not so flexible that it flaps around in the wind and generates lots of false positives.