Bridge between SoT and Agimus

• The visual tags should be declared and added to URDF / SRDF. Agimus-sot could parse them so that, for each gripper and handle, it is possible to know whether, there is a visual feedback. A possible implementation:
  • URDF: add the tag frame as a link.
  • SRDF: add a XML tag visual_tag that specifies the link. The issue (although nothing too hard) is that one must both parse the URDF and SRDF to know which joints have a visual tag.