Implementation

NaviBot Systems is working on building a navigation module for Dr. Michael Leverington’s Thirty Gallon Robot (a.k.a. Robot Assist Tours / R.A.T.). This robot will ultimately be able to give tours of the Engineering building here on campus.

In order to fulfill Dr. Leverington’s idea of a tour-guiding robot, NaviBot Systems will be building upon last year’s work to bring a navigation module to the Thirty-Gallon Robot. This module will require us to provide a way for R.A.T. to obtain a map of the Engineering Building, some kind of localization method to help R.A.T. determine its location, a GUI that tracks R.A.T.’s movements and status, and we need a module for reading and interpreting sensor data. Here are the ways we plan to fulfill these needs:

• Mapping with Gmapping Library. The gmapping library will allow R.A.T. to roam around Engineering Building and construct its own map using sensor data.
• Sensor Data via XBox Kinect. R.A.T. will use an Xbox Kinect to acquire its sensor data. The data from the Kinect can be converted to 2D sensor data to use during the 2D map construction.
• WiFi Localization. R.A.T. will use NAU’s WiFi network to ping nearby routers and determine its relative distance using signal strength.
• Navigation with Map, Localization and Sensor Data. R.A.T. will use all of the above mentioned techniques along with wheel counters to measure distance in order to traverse through the building.
• Graphical User Interface. We will be developing a GUI using Qt Designer that tracks R.A.T.’s movements and status as it traverses the Engineering Building.