[Dprglist] DPRG Annual Meeting Minutes and Rules for new competiton

[Doug Paradis]

John,
Yes, I think the judges would consider a pulverizing robot bad behavior or
unsafe <grin>. However, the rules do not require that you have to use a
vision system or the visual light spectrum.

I suspect that the issue boils down to resolution. It is very easy to have
a high resolution camera, but much more difficult to have a high resolution
radar, ultrasonic, or similar system. The resolution of the camera stems
from the fact that it is a sensor array, not a single "pixel" sensor.
Sensor arrays using other methods then visual or IR or UV spectrum light,
are hard for the hobbyist to obtain or afford. This ($$$) is the appeal of
vision systems. Also, there is a lot of exciting research and tools being
developed currently for vision systems that have a lot of interest.. I do
not think that it is just anthropomorphism.

Of the single "pixel" sensor approaches,  I think that LiDAR (note: LiDAR
arrays are now starting to show up) could be used by varying the scan
potential targets in an x and  z directions (assuming y is forward) as the
robot approached a potential target. but the system would be more complex,
but maybe not much more expensive.  As for just using the visual spectrum,
I know that Ray Casler is using IR images and laser illumination on some of
his vision systems.

The rules allow for decoy objects, objects that might trick the robot into
wasting time while collecting items. The idea is that the robot can't just
"sweep the floor", unless it is very fast.

The rules also do not say that the robot has to retrieve each object
separately or unload the objects. A robot could just collect the objects,
placing them in a hopper on the robot, then drive to the home base. I am
sure that others will find ways to interpret the rules to come up with
different approaches.

I also could see the use of ultrasonic, radar,  or LiDAR  to speed the
location of potential targets, allowing the robot to know that not only
does it think it sees a target object in the direction the robot is
pointed, but there is also indications from other sensors that there is an
obstacle in that direction. The combined distance information from both
sensors could also confirm object size. This would help differentiate
between a baseball and soccer ball (or to be evil, a softball).

I think the new contest is an exciting challenge and I hope to field a
robot in November. It will force me to learn new skills and approaches. I
think the same is true for other club members. I hope you give it a try, I
always enjoy your robots. Well, maybe not the sonic ones... <grin>

Regards,
Doug P.






On Thu, Jan 17, 2019 at 1:33 AM John Swindle via DPRGlist <
dprglist at lists.dprg.org> wrote:

> Doug,
>
> So, I assume the judges would consider it bad behavior if the robot
> pulverized everything on the floor and delivered all the debris to the home
> space. A hammer and vacuum miss the intent? The floor would be clean.
>
> Obviously, I think vision systems miss the whole point because
> anthropomorphism (visible light vision) forces the robot to be limited to
> what people can do, to the extraordinarily popular notion that vision is
> the thing to do, while sonar, IR, radar, LiDAR, UV and beyond are available
> and work much better than vision. We perceive what we do because we are not
> as good as butterflies (vision) and dogs (smell) and bats (hearing). We
> should not limit our machines to our limits, but that is what we do.
>
> Just grousing because I haven't come up with a way to do it with cheap
> speakers and a $10 toy robot. "If it can't be done with sound, it isn't
> worth doing." Working on that.
>
> Thanks for hosting another rousing annual kick-off.
>
> Best to you and the group.
> John Swindle
>
>
> _______________________________________________
> DPRGlist mailing list
> DPRGlist at lists.dprg.org
> http://lists.dprg.org/listinfo.cgi/dprglist-dprg.org
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.dprg.org/pipermail/dprglist-dprg.org/attachments/20190117/770cae1d/attachment-0001.html>