Develop control algorithms in Ros

edit

In my experience, it depends on how much you want to achieve. The key point is at simulation step. If you only try with simple dynamic model of the system (few equations), Matlab is prefered. If you want to try advanced simulation (Gazebo, V-Rep) with many complicationted robots and sensors, ROS is prefered. I do not say we can not do the same thing with Matlab but it will take more time and efforts. Remember that, ROS is produced after Matlab. It means that the inventors in the early day had seen some problems with Matlab.

I agree with Adolfo: it ultimately depends on your definition of ease. For me however (as an engineering, not IT-student) I would say Matlab is easier because my knowledge of C++/Python is limited and I have always worked with m-code. I'm happy I stayed away from Matlab for my current project though where I'm sacrificing development speed for run time speed. I think it's worth the effort as I probably won't always have Matlab at my disposal.

Can I develop a control algorithms in Ros?

Sure, you can definitely develop control algorithms in ROS.

it is easier to develop in Matlab, but why? what are the differences?

Ease of development depends a lot on what you want to do, how much existing work you can leverage (and don't have to implement yourself) and the workflow you're most comfortable with. So, it cannot be generally said that ROS is easier than Matlab, or the other way around. In fact, Matlab and ROS are not exclusive, as the former has ROS bindings, which allows you (up to a certain extent) to combine the two in a way that suits you best.