Skip to main content
Interactive Robotics Laboratory
Yu Gu, Professor
College of Engineering and Mineral Resources
WVU Home
  • Home
  • Research
  • SMART: Shared Mobile Autonomous Robot for Teaching

SMART: Shared Mobile Autonomous Robot for Teaching

Sponsors: Benjamin M. Statler College of Engineering and Mineral Resources, NASA WV Space Grant Consortium, MAE Department, LCSEE Department and MathWorks. We also gratefully acknowledge the generous support provided by Analog Devices, iRobot and NetBurner.

Introduction

A mobile robot is a valuable tool for teaching students sensors, controls, programing, and of course, robotics. There is surprisingly very few choices of mobile robotics platforms on the market that is suitable for undergraduate teaching. In our opinion, an ideal teaching robot should meet the following criteria:

  1. Have a simple, low-cost, and rugged design yet sophisticated enough to accommodate for state-of-the-art algorithms;
  2. Easy to understand and program by entry-level students;
  3. Allow reconfiguration and expansion with modular components;
  4. Well supported by the community.

To meet these goals, we are developing a Shared Mobile Autonomous Robot For Teaching (SMART) building upon earlier progresses in the robotics community. This robot will be designed in a full open source fashion, which means the complete hardware design, software, and lecture notes will be shared online.

The SMART Robot

The envisioned SMART design is shown on the left. The robot builds on an iRobot Create motion platform. The processing power is provided by an ASUS X202E laptop (other small laptop with similar configuration would work as well). In addition to sensors that are already in the Create, other on-board navigation sensors include a Microsoft Kinect, six Maxbotix LV-EZ2 ultrasonic rangefinders, and an Analog Devices ADIS16405 Inertial Measurement Unit (IMU), which include a triaxal gyroscope, a triaxial accelerometer, and a triaxial magnetometer. The IMUs and rangefinder data are recorded through a Netburner Mod5213 microprocessor before sending to the laptop with a serial connection. The main software running on the laptop include MATLAB 2013a and Microsoft Kinect SDK V1.7. The selection of MATLAB as the main programing environment will make programing easier for undergraduate students, especially ones with Mechanical and Aerospace Engineering background. The interface between MATLAB and iRobot Create is based on the toolbox developed by Joel Esposito and Owen Barton at the United States Naval Academy. The Create simulator is based on the one developed by Cameron Salzberger, K-Y Daisy Fan, and Hadas Kress-Gazit at Cornell University.

Development

The development of SMART started from May 16, 2013. The MATLAB-Kinect interface and MATLAB-Create Interface were tested. Mechanical design is mostly finished. A first prototype is printed with a 3D printer:

3/4 side view of 3-D printed prototype   Bottom of 3-D printed prototype


The PCB design is also mostly finished with the first board tested:

Mock-up of board design

Prototype of circuit board.


The completed SMARTrobot:

Mostly back view of completed SMARTrobot.     Mostly front view of the completed SMART robot.


Seven SMART robots have been successfully used in the classroom for almost a semester now!

A full set of  lecture material is now posted.

Students using the SMART robots in a lab setting.

We will post additional updates as the development continues. The robot design files and source code will be uploaded soon!

Online Resources