Rover configuration

Help & Support Apps & Software
1

Hi,

A bit quiet here atm, are people using something other than Roborealm at the moment?

I have a couple of questions.

As per usual, I see a couple of different ways of using the parts available to me.

As a test platform, I want a the rover to work in a follow me config, for example, following someone in a red t-shirt whilst avoiding obstacles. I have been playing with Roborealm for a short time and now want to get the hardware sorted before diving further into it.

I have at my disposal

A laptop system board stripped down to bare minimum for on-board control (will use desktop for initial bench testing)
SSC-32 to control pan and tilt
Sabertooth 2x10
Homemade max232 converter (if I connect Sabertooth direct to PC)
Bumper switches
4 x sharp IR sensors for cliff detection
1 x Maxsonar mounted on a servo geared to provide 360 sweep
4 x bumper switches
USB camera on servo pan and tilt
Chassis (tank style)

I also have a Make controller on the way, but that’s another story !

As per usual, I can see several possible methods of achieveing the goal, but wonder which you all think is best.

I want to start testing by interfacing the 2x10 to Roborealm on my desktop PC. Would this be best connected directly to the PC via a max 232 level shifter, or is there any advantage to connecting the 2x10 via 2 servo outputs on the SSC-32?

I am also learning C# express edition with a view to having Roborealm handle the vision, controlling the pan/tilt, driving the motors and C# handling the obstacle avoidance.

Any suggestions would be welcome as I am able to swap round the hardware into any configuration at this point.

Thanks

2

As far as vision goes, RoboRealm is your best choice in my opinion. Why? because its free :laughing: It’s really powerful and can be configured to do lots of things.

The RED T-shirt idea is an easy task but tracking red while still navigating around objects is more challenging. You could use a combination of IR and Sonar to navigate the floor and roborealm for the color tracking and get what you are looking for. Programming it all in to a seamless package is, for me personally, a programming challenge. I’m not the greatest when it comes to programming lol.

If I were to to this, I would start with object avoidance first and shelf RoboRealm for now and once you got that part out of the way, you can add in RoboRealm later. Another option is to write Pseudo code and work out your idea so you can code both ground nav & RoboRealm simultaneously if you wish. There are so many ways to solve this problem and my suggestion is just one quick idea. You may find out as you go one idea can lead to a dead end and a new approach is needed. Trial and error but persistence pays off.