Hey guys,
I had a quick question around mounting a LiDAR on a stabilised platform on a drone to counteract the pitch/tilt of the airframe. I was thinking about a steward platform with some servos however I'm using something like this below and it already sits 5" above the airframe. Would anyone have any idea? I'm planning on flying around 15-20mph and theres definitely going to be some pitch, if the drone comes near an obstacle but can't spot it because if the pitch, then its pretty much screwed, I'd like to counteract this by having it always horizontal.
In any other case I'd love to make something like the Scanse LiDAR however I can't really work out what motor they're using. I've tried looking for hollow stepper motors as I presumed they were using that, however they're all around 5mm in diameter and no slip ring will fit in those... Would anyone be able to shed light on this too ? :D
In other details I'm using the SF30/b by LightAware.
Thanks for any help, really appreciate it!
Replies
Hi Neil! Any progress? I would suggest a 2-axis gimbal (pitch &roll). But one more question; Why would you put two LidarLight sensors on your "scanner"? That might be good for robots that operates in a rather confined space like rooms or so, but for a drone... well I would rather put it in a "swinging" mode (let's say 120deg) forward facing. The response from your scanner would be faster since i doesn't have to go to a dead angle (the part behind your drone). And one more thing about LidarLight and similar sensors. They operate in a line fashion. The true lidars are scanning an area and this is a HUGE difference.
Let me know if I can be of any assistance. (I'm designing custom gimbals btw)
From what I've heard, gimbals may be too slow for it. I'll PM you since you make custom gimbals and we could do with some of that.
We're not using two LiDAR lite scanners, just the one SF30/b. We've decided to go ahead and order then SF40/c LiDAR. We want to do SLAM, hence why we don't really want to use a 120 deg LiDAR, we would of gone with those hokuyo lidars.
Yep, the true Lidar scanners are a lot more expensive, like the Velodyne products. I'll drop you a PM shortly :)
Opss! Sorry, from the picture you post ...it looks like two LidarLight opposing each other.
Please check this two videos below. Handheld! So you bought the SF40/C. Well neat little thing. Just a bit too pricey but for 100m that should be OK. Concerning gimbals it shouldn't be a problem since a speed of light is far faster than any movement a drone might do :-). And gimbals are quite good to "catch" that movements.
https://www.youtube.com/watch?v=F8pdObV_df4&list=PL0E462904E5D3...
https://www.youtube.com/watch?v=sieBqVxTz2c
It just pulled me in :-). OK, so her is an example of what might a gimbal for SF40/C look like. But I'm still not sure if you need one. Because, as I said before; light and sampling rate are quite fast. Anyway there is a lot to discuss here. It really depends on the application you are trying to achieve.
for a sake of simplicity I didn't put wires, electronics, bolts... :-)
@Dorjano - Very nice!
Which motor driver and IMU are you using or do have your own?
We've got some ultra-long range laser altimeters (non-scanning) coming out later this year that could use a very small gimbal to keep them aiming vertically downwards. Total weight is in the region of 40g. We can modify the housing to make it fit straight onto one of the motors if necessary. Ideally it should have at least 45 degrees movement in two axes.
This gimbal was designed as a proof of concept last night and is not done yet. Anyway the used components would be:
BaseCam SimpleBGC 32-bit Tiny
(I use BaseCam gimbal stabilizers mostly. This one is just for two axes and is tiny. But for two axes PixHawk is equaly OK and you don't need addidional electronics)
dys Gimbal Brushless Motor
(In this instance I'm using BGM2608-70T-8.5 as I do have them in my CAD library already. Plus for 270gr as the sensor weights this motors are perfect - maybe a bit to big. Ah, yes ant they have a nice hole for a slipring if you need it.)
LioghtWare SF40
(The sensor here is SF40, but I have the impression that you know it better than me :-)
The gimbal itself has a movement of +/-45 deg in Roll and +/-60 deg in Pitch direction. For a drone this is an overkill since at that angles means that it is stalling and falling down. Unless you have a 3D racing drone :-).
The total weight for the sensor is 40gr or the whole setup? I will assume this is for the sensor only. Well you can make a very nice tiny gimbals for this. There are some small gimbal motors available or you can go in extremes and make just a arch motor (this one is too big,but to get you an example).
D.
p.s.
LaserDeveloper are you form where I think you are?
Wow dude that looks great! I've dropped you a PM in regards to that. Are you 3D printing these things?
In regards to the gimbal, the reason why we want one because of the fact that our drone is capable of travelling up to 60MPH, at that speed it tilts downwards significantly thus if the LiDAR was stuck to the frame it won't be able to see obstacles in that the drone may hit.
LaserDeveloper is LightAware :-)
Therefore, yap, you need a gimbal. The one (designed yesterday night) on the pictures is based on real SF40 with some assumptions. So centre of gravity is more or less there. Since you are looking for a custom gimabl t can be fully optimized to be lightweight and super responsive, not like a gimbal for cameras. Right now I'm building a gimbal for a camera with exchangeable lenses and it is a nightmare to make it universal-ish.
I'll post 3D files for everyone to use it after it's done.
Ahh forgot! Yes I do have a 3D printer so I'm printing this thing out. Plus I have a CNC 3D router. You have to have this things if you want be a developer, otherwise outsourcing might kill your dev-ambitions.
Check out this DIY project:
DIY 360 degree realtime outdoor LIDAR
Looks like they use it on up to 10Hz rotation scan rate.
However Lidar Lite sensor availability is still a big question.
-
1
-
2
of 2 Next