IR-LOCK Sensor for Precision Landing, etc

We finally have a progress report on the integration of the IR-LOCK (Infrared Pixy) vision sensor with Pixhawk. A group of students at the Georgia Institute of Technology is using the sensor for their 'Package Delivery Drone', which requires precision landings. They plan to use GPS to travel to a waypoint, and then use the IR-LOCK sensor to land precisely on an infrared beacon. Recently, they were able to get an IRIS+IR-LOCK to auto-hover over an IR beacon (see videos).

Other applications for IR-LOCK/Pixhawk are in the works (or at least in mind):

Search&Rescue (auto-search for emergency IR beacons)
Moving target following
Aerial surveying (I will be able to post a relevant video soon)
... your thoughts/ideas are welcome :)

We are in the process of making the IR-LOCK/Pixhawk interface more developer-friendly. The current developments (by the student group) are based on an IR-LOCK sensor 'driver' and codebase discussed on this FAQ page (link). A 'more official' implementation should be under development soon. The relevant discussion is here (link).

Small IR 'Pods' can be detected by the IR-LOCK sensor at about ~30-60ft. The IR-LOCK sensor reports the (xy) position of IR targets to Pixhawk at 50Hz.

This is awesome!! I had a couple of ideas (although I am sure you have thought of these):

1) Could your "IR-Lock mode" function in addition to GPS lock in some way? Scenario --- I may place an IR Beacon at a rally point or at the end of an Auto Mission... Drone would navigate to the location where beacon is expected to be and position hold using GPS UNLESS, IR Beacon is spotted. --- I suppose you could call this a failsafe to GPS.

2) Would the Raspberry Pi camera suffice for this application? --- This would be most convenient as the cameras are cheap, interface to a device which can do other things as well... e.g. LTE telemetry...

http://www.dx.com/p/waveshare-zooming-night-vision-camera-board-for...

The camera is almost good enough for FPV in the dark with IR spotlights

3) WRT IR camera positioning... Brushless gimbals definatley have a degree of drift to them but a simple 2 axis gimbal with little slop in it I am certain will be able to bring the IR camera back to NADIR orientation (or at least the same position as every time). The ability to set the position of 2 channels when initiating "IR-Lock Mode" would be an awesome addition :)

Keep it up. Much enthusiasm from this side indeed!

You need to be a member of diydrones to add comments!

Join diydrones

Comments

Dylan Eave February 23, 2015 at 10:50am

Soon as you are ready with a stable build I am happy to post a video for testing of number 3 if you would include the ability to override two gimbal channels...
Dylan Eave February 23, 2015 at 10:41am

This is awesome!! I had a couple of ideas (although I am sure you have thought of these):

1) Could your "IR-Lock mode" function in addition to GPS lock in some way? Scenario --- I may place an IR Beacon at a rally point or at the end of an Auto Mission... Drone would navigate to the location where beacon is expected to be and position hold using GPS UNLESS, IR Beacon is spotted. --- I suppose you could call this a failsafe to GPS.

2) Would the Raspberry Pi camera suffice for this application? --- This would be most convenient as the cameras are cheap, interface to a device which can do other things as well... e.g. LTE telemetry...

http://www.dx.com/p/waveshare-zooming-night-vision-camera-board-for...

The camera is almost good enough for FPV in the dark with IR spotlights

3) WRT IR camera positioning... Brushless gimbals definatley have a degree of drift to them but a simple 2 axis gimbal with little slop in it I am certain will be able to bring the IR camera back to NADIR orientation (or at least the same position as every time). The ability to set the position of 2 channels when initiating "IR-Lock Mode" would be an awesome addition :)

Keep it up. Much enthusiasm from this side indeed!
Thomas Stone February 23, 2015 at 8:46am

Our latest test results are in this video.
Thomas Stone February 20, 2015 at 8:37pm

@Dylan

Thanks for the proposal! The precision landing controller is heavily dependent on the angle of orientation of the IR-LOCK camera, so we like to have it fixed to the orientation of the quadcopter (at least for now). We probably need to get this setup working reliably before we move on to more features.
Thomas Stone February 20, 2015 at 8:27pm

@Crispin

Actually, that was sort-of the original motivation (IR-LOCK), but now we are primarily focused on the precision landing objective. It is probably better that we start with a stationary target as we begin the official integration with Pixhawk/ArduPilot.... If everything goes smoothly, I assume there could be derivative applications. But I also look forward to the future of optical object tracking.
Thomas Stone February 20, 2015 at 8:15pm

@Glen

Thanks for sharing! I also like the idea of passive targets, but we have had some good success with small active targets (IR beacons/pods). I hope to bring more good news in the future.
Dylan Eave February 19, 2015 at 10:29pm

Another user very excited about this project!

May I propose that, in the interests of maximizing the utility of this technology, AC3.3 allow for both gimbal based IR Beacon tracking AND precision landing?

>> Perhaps when entering LAND mode the gimbal points to NADIR position and PIXY begins precision guided landing...
Crispin February 19, 2015 at 9:52pm

Could something like this be used as a precursor to the all singing all dancing Follow-me? The days of optical recognition object tracking will come soon (cannot wait!) but this seems like it could also be a first step?
Glen R. Sitz February 19, 2015 at 9:37pm

From a "Newbee",

This project is of great interest to me. I am new to flying an Iris+ (and drones in general). I fly from a fairly challenging environment. I am on the side of a very steep hill, there are very few and very small level places, there are lots of trees and the major crop here is surface rocks. So precision landing is very important to my drone.

My original thought on the subject was to take a 2 meter square bit of concrete and paint a target on it and see if I could get the Iris to land there. I was just beginning to consider passive sensors when this topic opened up.

The idea of active sensors on the landing zone is very interesting.

I will follow this thread with great interest.
Thomas Stone February 19, 2015 at 8:40pm

@Crispin

Great! Thanks for your business. Will ship tomorrow (Friday).

of 6

This reply was deleted.