Re: Ultrasonic sensor interactions

At 06:22 PM 5/21/2006, steve wrote:
>Brian Davis wrote:
>
>>    On a related note, I'm trying to see just how much interference 
>> two US sensor
>>working in the same room actually produce. The result seems to be 
>>"darn little",
>>but I need to test some more. This may not be as big a problem as 
>>we're worried
>>about, I'm unsure now.
>
>Within the same room - is unlikely to be a problem - but think about two
>Sumo bots trying to find each other or two robots doing some kind of
>cooperative tasks in a contest or something...not likely to be so
>easy I think.

Agreed... perhaps surprisingly, based on dozens of hours of testing, 
with 1) sumo bots with a ring of sonars on all sides 2) Lego bots 
with two sonars on the front, we found no discernable problem. This 
is an anecdotal report, but the first-order answer to the question 
seems to be "it's nowhere near the problem you might expect", and 
typically appears to be no problem at all.

Since we're working on a long term SonarMotion project here: 
https://sonarmotion.dev.java.net/ we plan to do some more deliberate 
testing with test equipment monitoring modules with deliberate 
worst-case timing and placement.

In practice, other problems dominate: 1) reflecting at a low angle 
causes the ping to not reflect back to the receiver at all, giving a 
false max-distance reading, often just before you scrape up against 
the object you were hoping to detect. 2) absorbent objects like 
drapes or fuzzy fabrics don't give a strong echo. 3) highly spherical 
objects don't have enough planar surface perpendicular to the sonar 
module. 4) sonars with a narrow beam aren't as good close up as those 
with wider or multi-lobed beams. For example, the Vex sonars won't 
see my hand edge-on 20-30 cm away if it's located midway between two 
modules offset by 15 degrees. But the Vex modules seem to have better 
long distance performance, at 2 meters and up. The SRF04 modules on 
the other hand have great detection up close, over a wide angle, and 
are not so useful at distances in excess of  a couple of meters.

That's our experience, anyway... we haven't used the new Lego sonars 
- we are waiting for the NXT kits like a lot of other folks - but 
expect them to be similar, assuming they are in the 40 KHz range with 
similar beam patterns.

Regards

Bruce



If the human brain were so simple that we could understand it, we 
would be so simple that we couldn't.  - Emerson Pugh, 1977