Re: [HBRobotics] Tennis Ball Robot

[Sergei Grichine]

After you perfect the art of picking up tennis balls - please, please come to my place to the next challenge  - pine cones. I'll put steak on the grill ;-)

Best Regards, 
-- Sergei Grichine
   

On Thu, Sep 1, 2022, 10:44 PM Chris Mayer <cmay...@gmail.com> wrote:

I just saw a cool video, then found others, for robots that collect tennis balls.

I'm just as lazy as the next guy, so I think this could be my next project.

As I have a full time job, my projects usually take many months to years, if they finish at all :)

Is anyone interested in forming a small group to design and build a cool tennis ball retriever?  Can't be that much harder than Tabletop challenge level III, right?

https://www.youtube.com/watch?v=tPjPdUyENeg

https://www.youtube.com/watch?v=Rg1AtV8wguU&t=119s

--
You received this message because you are subscribed to the Google Groups "HomeBrew Robotics Club" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hbrobotics+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/hbrobotics/ccf20969-a5f6-4003-9d2a-89c1afb6b3ddn%40googlegroups.com.

[thomas...@gmail.com]

Ah, that would be a great project!

To view this discussion on the web visit https://groups.google.com/d/msgid/hbrobotics/CA%2BKVXVNDGr%2BaJM3LmQYJuVcLeCApxzLNV1mm71k81%3Dd%2B9OuJRQ%40mail.gmail.com.

[Steve " 'dillo" Okay]

On Thursday, September 1, 2022 at 8:55:07 PM UTC-7 vital...@gmail.com wrote:

After you perfect the art of picking up tennis balls - please, please come to my place to the next challenge  - pine cones. I'll put steak on the grill ;-)

Best Regards, 
-- Sergei Grichine

This is probably marketable as a fuel-reduction/mitigation grounds-clearing robot. It would need to be a bit ruggedized and/or on a suspension to get over the ground.

Some sort of sweeper mechanism would probably do the trick. Now this makes me want to give those "electric wheelbarrow" motors somebody linked to on Ali a second look.

That plus a VPD like an OAK/OAK-D would get you well on your way.

'dillo

[Marco Walther]

Way too complicated;-) Turn off the blade on your auto-lawnmower and
attach a rake ;-) Something like
https://img.ricardostatic.ch/t_1000x750/pl/1084326240/0/3/ ;-)

-- Marco

> 'dillo

>
> --
> You received this message because you are subscribed to the Google
> Groups "HomeBrew Robotics Club" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to hbrobotics+...@googlegroups.com

> <mailto:hbrobotics+...@googlegroups.com>.

> To view this discussion on the web visit

> https://groups.google.com/d/msgid/hbrobotics/77ef07b9-06a8-431e-a382-6199ca510d9bn%40googlegroups.com
> <https://groups.google.com/d/msgid/hbrobotics/77ef07b9-06a8-431e-a382-6199ca510d9bn%40googlegroups.com?utm_medium=email&utm_source=footer>.

[Chris Albertson]

I've been thinking that a trash pick up robot would be a good project.   It would be easier if you defined just one object as "trash".   I had thought of using red Coke cans but bright yellow tennis balls would make it much easier.  Balls are easy to grip because they are the same from all directions.  Orientation does not matter.

How hard is it?

Let's assume we have a "correct size" mobile base that responds to ROS "twist" messages.  This assumption is easy to do and gets you 3/4 of the way there.  You need motors and wheels and encoders and battery and so on.  You would likely need four wheels.   

Next you need a tennis ball locator camera and software that will find a ball in a photo.

Then a "gripper that can pick up a ball.  I'd say to use a vacuum hose.  Fingers are too hard. The hose might be 2/3 the ball diameter.  The arm that moves the gripper only needs 3-DOF as orientation does not matter for balls.  But vacuum does not later generalize to trash pickup.  It only works for balls.

You can do all the work on the arm, gripper, and camera on a tabletop with no mobile base.

The other way is to have a "dumb" robot randomly move around with some kind of scoop to catches tennis balls.  Given enough hour and some luck it would eventually find all the balls

To view this discussion on the web visit https://groups.google.com/d/msgid/hbrobotics/CA%2BKVXVNDGr%2BaJM3LmQYJuVcLeCApxzLNV1mm71k81%3Dd%2B9OuJRQ%40mail.gmail.com.

--

Chris Albertson
Redondo Beach, California

[Ralph Hipps]

that looks like a fun project!!

Agree the Oak D-lite cam would help out a lot here.

I wonder about the collection mechanism, would a rake work in grass? or do you need an arm & gripper?

tennis balls on a hard court would be easier, but still not trivial. At least you could collect them without getting caught up in the grass.

something like this:

https://www.youtube.com/watch?v=CRI4gUFm-sI

On Thursday, September 1, 2022 at 8:55:07 PM UTC-7 vital...@gmail.com wrote:

[Chris Albertson]

A tennis ball collector would be a good step toward building a more general-purpose trash collector.  So my preference would be to build an arm and gripper that eventually could work for all kinds of objects.  I'd want a 4-dof arm and detachable gripper.    I would use a 3D printed "Yale Hand" as the gripper.

But if all you ever want to pick up is tennis balls, try a vacuum device.  A tube that is just a bit smaller in diameter than the ball connected to a vacuum cleaner motor.  Then if you get the tube within an inch or so of the ball, the ball sticks to the end of the tube.

Even simpler but more specialized would be a simple bulldozer.  A "U" shape dozer-blade in front and then you push the ball on the ground to a collection area.

With enough budget, the most fun solution is to buy Spot from Boston Dynamics for $70k and let Spot fetch the ball and carry it back in it's mouth/arm thing.

In general the more specialized the robot the lower the cost.  If it only has to return tennis balls on a hard surface then a bulldozer is the lowest cost.  Adding more capability like onboard board ball storage or the ability to work on grass adds a lot of cost.   I think the best first set is to nail-down requirements or you can run the price way up.

To view this discussion on the web visit https://groups.google.com/d/msgid/hbrobotics/844f8e4f-b76c-42cf-aaab-d787934bfbd8n%40googlegroups.com.

[Dave Everett]

If you can find video the AAAI competition one year was collecting tennis balls on a court. There was a very professional entry, it collected all balls, and also one using Newton labs vision system, that will tell you how long ago this was.

 

I had it on VHS somewhere.

Dave

To view this discussion on the web visit https://groups.google.com/d/msgid/hbrobotics/CABbxVHu6XKqGAcb_SzWR5zzo80KzwBGwMv2PNoLQe6DEgxPhmw%40mail.gmail.com.

[Dave Everett]

This project interests me even though I don’t play tennis or even own a tennis ball.

I can imagine a product where the robot raced out to get a ball once it was stationary or moving only at ground level. The robot could push it up a tube to present it to the player.

Dave