ROS-Industrial is a industry-university consortium managed by Southwest Research Institute. ROS-I is "an open-source project that extends the advanced capabilities of the Robot Operating System (ROS) software to new industrial applications." ROS-I issued the following release on the matter.
Headed by Dr. Mitch Pryor, the Nuclear Robotics Group (NRG) is an interdisciplinary research group associated with the Nuclear Engineering Teaching Laboratory (NETL) at the University of Texas at Austin. NRG makes use of industrial automation hardware as it conducts graduate-level research targeting the energy sector. <Add something about NRG’s relationship with LANL?> As it has used a variety of C++ based middleware for its research in the past, it made sense for NRG to begin using and contributing to ROS-Industrial over the past year as a member of the Consortium.
The ROS-Industrial team at SwRI enjoyed working with NRG researcher Dr. Brian O’Niel who spent summer 2012 developing a 3D object classifier that was used for the ROS-I Automate demo (video link). Dr. O’Niel’s work demonstrated how quickly academic research to transition to practical use on real industrial hardware. In a period of a few months, his idea was in practice on a heterogeneous dual manipulator system that demonstrated many of the core capabilities of ROS-Industrial.
NRG has recently released a Multiscale Teleoperation Demo video (below) that shows a natural human interface to an industrial robot. In the video, Ph.D. student Jack Thompson uses hand and arm motions to set waypoints for a simulated Motoman manipulator. A Kinect 3D sensor observes Jack’s motions, and then his ROS/PCL-based software nodes interpret the motions and convert them to tool poses. What is unique: Jack has a separate input control that scales the system’s sensitivity to his hand/arm motions. If he wants the robot to execute a small/delicate motion or a large macro-motion he is able to do so by scaling the sensitivity accordingly. We look forward to his next accomplishments as he begins using the capability with NRG’s Motoman SIA5 robot.