Let's Make Robots!

Lets Make Robots or Lets Make Radars?

A while back I notice the simularity between the Ultrasound/IR sensors we use on robots for object avoidance to the concept of basic RADAR. (This is more like UDAR or IRDAR). Direction and range information are what the controller uses to detect objects (RADARs basically use the same information.)

So instead of building robots for a while I decided to see how good I could make my own IRDAR. Wow! One can adjust trigger point, gain and a bunch of other factors and have a way-cool little system!

For processing, I am first using Uno32 running the standard Arduino sketch. The 32 bit PIC32 on the Uno32 really speeds things up like math, averaging and anything else needing processing power. The display is done on a PC using the free Processing program. (The programming is VERY close to Arduino Sketches.) A lot of the great display section has been "adoped" from the very smart "luckylarry". (His software is old and I do not plan on selling any of it, credit will go to luckylarry at the top of each proogram.) With my many modifications and changes the resemblence to the original programs is minimal. He had a great place to start, but so much more is possible (with my hardware setup basic RADAR funtionality exists at the Uno32, minus the display).

My final plans will be to include a USB based Nerf Missile Launcher so when someone comes in the room the system will "lock-on" to them and fire a foam tipped missle or two. This has been done on the web with a Ultrasound ranger, a stepper motor rotating the sensor and a PIC chip feeding the information to a PC. I LOVE their display and wish I could somehow adapt it to my system (but it is written C# and I am not THAT skilled with that language). Maybe with much more work...

Who gets the picture(s), schematic, and software when I am done? LMR does!!!

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.

http://letsmakerobots.com/node/32599

my project using IR to make a 3d image, well kinda ,i have some code up there for the Arduino and Processing it's not exactly lidar (light radar ) might help :)

Ok, so you are talking about making an IRDAR. The software and processor are not important. The hardware is critical. What hardware are you using?

Are you talking about building your own IR rangefinder or are you just talking about mounting a Sharp IR sensor on a servo? Gareth did a great job of this using a LCD graphics display a few years ago.

I went back was just reading about some of the work you have done with the Dagu Compound Eye and it was indeed both fascinating and informative. Your work with the sensitivity and the amplifier was well worth storing for future reference.

One thing I like about the Sharp IR Rangers is their insensitivity to external light fluctuations.That helps makes the design of the system somewhat easier. What I dislike is IR tends to ignore certain materials. (Funny, to avoid a Nerf missile system all one would need to do is to surround oneself with a material the IR would ignore.) One possible solution I have considered is also using a Ultrasound module that would pick up such materials (ignoring its own set of materials). Processing all this into usable information might prove to be a challenge.

I love your "dog" and other robots using the compound eyes reaction to the presence of a hand. Very impressive indeed!

 

IR, Ultrasound and even the human eye all have things they can't see. Ignoring our own eye's limit to see infrared or ultraviolet just think of camoflage. This is why a robot really needs at least 2 different types of rangefinder or vision sensor.

Now I'm back in China I've had a chance to dig through a huge mountain of Gareth's projects. Here is his LoMoR.
http://letsmakerobots.com/node/11259

I have seen some other attempts on LMR to create 3D images by scanning the rangfinder left to right, top to bottom. the result was slow but impressive. Unfortunately we did not have the "collect" button then however I did find this example:
http://letsmakerobots.com/node/10883

 

That was great information and it will go a LONG way in helping me build the best "IRUSDAR" I can build.

I was going to "swing" the IR detector from a tiny blue servo, but adding the ultrasound unit means I better switch to a full sized servo. (The blue serco probably would work for a bit, but the gears would wear out quickly.) Then it just becomes a mechanical design detail of making the whole thing rotate smoothly 180 degrees.

Finally the software would have to process the analog and pulse data and "really" see what the radar was seeing (that shouldn't be really that hard).  The display sofyware may not change, but the result would be a more detailed and accurate final diaplay. (The distance could also be more accurate becacuse both sensors have better accuracies at different distances.)

The good news is the company will take the board back with a RMA number and see if the chip was electrically damaged or the part had just mechanically failed. If it was a mechanical failure I get a new board. If it was electricaly damage beyond static charge I just get my old board back. I think that is totally fair.

By the way, have a happy and safe 4th of July!

I am using a Sharp Long Range IR Ranger and it is working even better than I expected. The range is excellent (it will easily cover a small personal space, up to 150cm) and the resolution (ability to pick out out individual objects) is also very good. Sharp's sensors really are an excellent feat of engineering. I am using a servo and getting about a 180 degree sweep. Azimuth (horizontal) seems sufficient to do what I need, but I would like to use an altitude (vertical) measurement to make a really interesting quasi-3D map of the scanned area. It would take more time to make a full single sweep, and thus be slower to see objects that enter the detection (and possible targeting) area. The results from a multi axis, multi detector setup (like the Dagu eye) would be fascinating! (Not totally unlike a multi antenna "phased" RADAR array.)

I have seen the results displayed on a LCD and also on a series of LEDs. Color information "adds" to what can be visualized so I think a PC running a program like "Processing" is of great benefit (but is not absolutely required).

For totally different hardware and a beautiful (C#) display see: http://www.steffenschuette.de/steffenschuette/elektronik/us2bRadar/  It also can aim and launch Nerf rockets, nice! (Hardware design information and the source code of the C# based display are available.) I have seen this setup functioning on YouTube an the results are nothing less than astounding!

See: http://luckylarry.co.uk/arduino-projects/arduino-processing-make-a-radar-screen-part-3-visualising-the-data-from-sharp-infrared-range-finder/  (Yow! what a huger URL!)  Some of the initial code I am using can be seen there.

This little setup in no way gives the fantastic results a microwave based "standard" RADAR gives and you can't expect it, but it is really fun to work with and shows what is possible with MUCH less expensive hardware.

DAMN Hong Kong airport WiFI...   Tripple post.

DAMN Hong Kong airport WiFI...   Tripple post.