Validation between simulation and the real world
Hi all,
this question is maybe more like an open or theoretical one.
As one part of my work with the robot simulation I have to make some tests in the real world as well as in my simulation to validate the usefullness of my work. We want to use the simulation for further development and tests before we try them on our real hardware.
Maybe you have some proposals for me, you have some experience with this topic yourself or know some points that I really shouldn't miss. For this purpose I plan to build a recreation of a simple floor or room in our university in Gazebo as the scope for these tests.
At this point of my work we don't really care about a completely correct physics model or that our pid parameters would work in reality, as long as we have a simulation that somehow behaves like our real robot (for it has e.g different engines for the joints). I was more thinking about something like letting the robots (real and simulation) drive to a defined position and look at the deviation, put an obstacle in their way and look at their behaviour etc.
Thanks for your help,
Christoph
So, what is your question? For Q&As in general specific questions usually yield a better result in terms of the number and quality of answers.
See Nate's comment. Thats the kind of things I had in mind.