We are a bunch of young mechanical engineering students with limited
Our term project is to build small robot(s) to collect pingpong (table
tennis) ball in an indoor environment. We can design and build fancy
mechanical devices but we are so good in electronics.
Would you share your toughts with us how we can locate the balls ?
(Please keep in mind our lack of experience inelectronics, microprocessors
It all depends on the details ...
There is probably a lot to be said for the simple 'blind mans buff'
approach - rush around with some touch feelers, and quickly grab a ball when
you touch it (perhaps a sort of 'mousetrap'). You would need some sort of
arena boundary detection, and would like to see the bot bouncing around in
such a way as to scan most of the arena. Because it could cover a lot of
ground, under some conditions this might even be the most 'efficient'
Otherwise you are into sensors to detect the balls remotely. You will
quickly discover that this is harder than it looks. If we discount vision
systems (eg CMUCAM), then the only realistic approaches are ultrasonics (eg
SRF08), or some sort of optical scan or rangefinder, looking for reflected
light. Both of these tend to work best at fairly short range, and become
confused near boundary walls, other bots etc - so you will still need to
'scan' the arena. If you want to make your own sensors, be prepared for a
significant amount of (interesting) development work and time.
Have fun !
I'd make up a 7.2v to 24v DC version or a shop vac, mount it on a robot
chassis, then run a simple methodical plot path on the course to vaccumm up
all the balls. Use IR sensors for walls to avoid touching them.
Ping pong balls are really light, the slighest breeze or touch c\will cause
them to move on you.
If you use one of those 7.2v RC ducted fan units, it should have sufficient
power to suck up the ping pong balls.