Replies to This Discussion

Permalink Reply by David Gitz on January 4, 2010 at 11:52am

I am working on a similar feature on a Quad-Rotor heli with 4 ultrasonics, though I am still in the conceptual phase and need to start working on the control system equations for such a thing.

• ▶ Reply

Permalink Reply by Stephen Buck on January 4, 2010 at 3:30pm

Same for me. At the moment, I'm just using a MaxBotix ultrasonic for the forward sensor. I plan to add two more MaxBotix sensors, one for the bottom (ground) and one for the top (indoor ceiling), and two Sharp infrared sensors for the left and right sides (walls). I need to do some testing to see which type of sensor (ultrasonic or infrared) works best for each direction. Both types of sensors are a bit expensive though - it would be nice to find a cheaper alternative.

Which directions are using the sensors on? Also, what range are you aiming for?

• ▶ Reply

Permalink Reply by Melih Karakelle on January 4, 2010 at 10:54pm

Hi Stephan,

I'm not suggesting this IR Sharp sensors, because it is affecting from colors. I was working with them for making contactless blind sticks, totally impossible, it is mesuring distances when you lucky, not everytime.

• ▶ Reply

Permalink Reply by Stephen Buck on January 5, 2010 at 5:38am

Good to know - so it looks like it's best to go with the ultrasonic sensors, even with all the problems they have. The main reason I was considering the IR Sharp sensors is that they are a bit less expensive than the MaxBotix sensors. A solution for autolanding would require a single sensor (bottom: ground) and a basic obstacle avoidance solution would require at least 5 sensors (top:ceiling, bottom:ground/floor, front:forward, left/right:walls). Even that would probably leave the corners exposed. Does anyone know of a good ultrasonic sensor, similar to MaxBotix, for a reasonable price? An open source solution would be really cool.

• ▶ Reply

Permalink Reply by Lucas Tan on September 12, 2012 at 11:58pm

Hi,

I am also doing related things like what you're considering. So if i were to follow what you're doing, i'll need at least 5 of the sensors? I need to know more about how it works. It is ok that you share this with me?

Thanks!

• ▶ Reply

Permalink Reply by Joe on January 4, 2010 at 4:32pm

A few years back some guys in our lab tried to use an ultrasonic sensor with a micropilot to do auto-landings. The problem that they ran into was that the aircraft was traveling fast enough that by the time the signal returned to the aircraft the sensor had already moved too far away to receive it. At least, that was what they diagnosed the problem to be. There are some ultrasonic sensors with larger receivers that, supposedly, have been successfully used for auto landing on other aircrafts. The bigger problem I would worry about is that grass is going to give you a messy return signal at best.

Infrared sensors have a similar problem where the varying reflectivity of different surfaces can make things effectively invisible.

Other alternatives include laser altimeter (very expensive), gps, and barometric altimeter. The Piccolo autopilot is able to do autonomous landing using gps and barometric altitudes. And I expect that it could do it using just the gps altitude.

• ▶ Reply

ModeratorPermalink Reply by Morli on January 4, 2010 at 11:09pm

GPS or atleast what is available to hobbyist now, the altitude estimation may not work for precise landing calculation and flaring at right point , laser does it but being expensive is right now out of hobbyist reach , so ultrasonic sensor &/or barometric sensor is best possible solution that needs more scrutiny and attention/work IMHO

• ▶ Reply

Permalink Reply by Melih Karakelle on January 4, 2010 at 11:10pm

Yes i'm afraiding about same problem,
- GPS altimeters not sensitive, i saw altitude errors around 7-10 meters.
- Barometric altimeters looks good only with short flight times. Barometric pressure changing everytime.
- Infrared useless affecting from color changes and reflections.
- Ultrasonic looking good option already, only problem is total flight time of acoustic wave, but i guess i can resolve this.
- Laser altimeters expensive because need good optics.
- Radar Altimeters best option since last 50 years. It can read real ground, cheaper construction than lasers but they are using different way(not flight time of wave) and i need much more time and rf test equipments for making this.

Ultrasonic sensors already best option,cheap, you can find it from everyware and every beam profile.

• ▶ Reply

ModeratorPermalink Reply by Morli on January 4, 2010 at 11:34pm

wide beam model (model 0) ultrasonic sensor might be good for issue mentioned by Joe or is it possible to use 2 sensor one as tx( near nose) and other as rx( in the tail) timed coordinated to and keeping the wiring harness symmetrical , angling/pointing the sensor little forward are some of the thoughts in my mind. it has been in my wish list for some time now , your work will help lot of enthusiasts. Thanks for sharing :-)

• ▶ Reply

Permalink Reply by Melih Karakelle on January 5, 2010 at 2:14am

Two sensors makes some installation problems etc. I'm thinking intalling 45 degree to forward.
It have 2 advantages;
1- little beam error,
2- Sensing ground shapes before.

I will use 7 meter range sensors for testing, i guess last 3 meter will be enough for smooth landing.

And i have a secret weapon for landing, it is very high technology sensor system but i need time for perfection. Now it is just a secret :)

• ▶ Reply

Permalink Reply by PavelEkb on January 28, 2010 at 9:43pm

I'll try to make UAV with autopilot landing using three sensors:
1) GPS - to proximitely (about 15 meters) find the land
2) Ultrasonic sensors - to precisely (1 inch) find the land from altitude about 6 meters
3) Pressure altimeter (about 15 cm) - to find the land second time

I'd find the land descending with ultrasonic sensor with normal UAV speed. Then I'll lift up and make second pass for landing with low speed knowing the height very precisely.

I think two passes for landing make it very carefull.

• ▶ Reply