Overview of the Segway RMP

Segway Robotic Mobility Platform (RMP) a version of the Segway Human Transport (HT) developed by Dean Kamen. Vanderbilt was chosen as one of only a handful of research groups across the nation to receive a RMP.

The other research groups are: University of Massachusetts, Carnegie Melon University, the University of Michigan, MIT, USC, Georgia Tech, University of Pennsylvania, Naval Research Laboratories, National Institute of Standards and Technology, SPAWAR Systems Center, and NASA.

The Segway Robotic Mobility Platform (RMP) is a version of the Segway Human Transport (HT). The RMP has been modified to function as a self-balancing robot. One of the main features of this project is that it allows for a robot to function in the types of environments people have designed for themselves. Due to its self-balancing nature the RMP can be taller than many other robots. Due to its two-wheeled design it also has approximately the same standing area as a person. With these two concepts in mind the RMP is the perfect robot for human environments.

Figure 1: CIS's Segway RMP

The RMP has a battery life of 2-3 hrs and can carry a payload in excess of 200 lbs. The RMP is robust and versatile and can travel virtually anywhere a person can travel.

The Segway RMP was distributed with the idea of unifying certain fields of robotic research by providing separate research groups, all with widely varying research interests and strengths, a completely uniform and interchangeable platform with which to develop their own particular research concepts. Basically what this means is that an RMP from any of the research institutions mentioned above can be shipped to any other institution and plugged into the new system and run without any other software or hardware work. Because of this each University can focus on its research strengths and separately apply them to the RMP.

Here at Vanderbilt University we have been working on form of Short Term Memory we call Sensory Egosphere (SES) where an objects position in reference to a robot is detected and posted to a geodesic dome surrounding the center of the robot. A stationary SES has already been applied to our humanoid robot ISAC.

Figure 2: ISAC Humanoid Robot

Our work on the Segway RMP is going to involve developing a mobile SES to place on the RMP and develop a form of SES sharing so that ISAC and the RMP can work together as a team to accomplish task.

A good example of this would be to have a layout of tools on the RMP with their positions and orientations noted. If ISAC were to require the use of one of these tools a command would be sent to the RMP and using the SES sharing the RMP could position itself in such a way to allow ISAC to appropriately grasp the necessary tool. The Segway could then leave ISAC’s workspace, carry out other task, and return once ISAC is finished with this tool and perhaps in need of another tool.

To demonstrate what we are doing we are initially working on a “Grasp-the-Barney” task involving both the Segway RMP and ISAC. This task will require for the RMP to know its position and orientation relative to ISAC, the position and orientation of the Barney it is carrying, and the position and orientation ISAC will require to grasp Barney. All of this is going to require that both robots are in constant communication so as to effectively and efficiently complete this task.

Figure 3: ISAC System Configuration