Exposing Middle School Students to Robotics and Engineering through Lego and Matlab

摘要

Abstract Exposing Middle School Students to Robotics and Engineering through Lego and MatlabUse of robotics in the K-12 environment is becoming increasingly common to introduce studentsto educational and career opportunities in engineering and technology fields. Moreover, severalrecent studies have shown the effectiveness of robotics-based, hands-on science and mathlessons vis-à-vis traditional classroom instructional practices. A wheeled mobile robot providesnot only a compelling hook to engage students’ interest, it also brings new meaning to classroommath and science concepts by connecting them to real-world applications.This paper considers a novel application of the Lego NXT platform and the RWTH-MindstormsToolbox for Matlab to create a series of classroom activities that expose and engage middleschool students in all aspects of controlling a differential-drive mobile robot. Lego bricks areused to construct a mobile robot consisting of two non-steerable, powered wheels, and one castorwheel. The mobile robot uses a bluetooth connection to transmit wheel rotation data to acomputer base-station running the Matlab program. An odometry model running on Matlab usesthe wheel rotation data to keep track of the rotation and displacement of the robot on a Cartesianplane. This hardware-software platform is used to create and implement several lessons thatillustrate real-world applications of concepts from middle school curriculum.In the first lesson, the relationship between wheel rotations, wheel circumference, and distancetraveled is derived. The effect of wheel diameter on forward velocity is assessed and the errorsthat may accumulate from the use of an incorrect diameter of the wheel are examined. Thestudents use their knowledge of wheel rotations versus displacement to measure the length andwidth of the classroom. Next, the notion of feedback control is introduced through a controllerthat enables the robot to move a commanded linear distance. In the second lesson, the effect ofcommanding one wheel to move slower than the other wheel is observed. This leads to thederivation of the relationship for turning radius and the velocities of the individual wheels of therobot. Next, a complete odometry model is studied wherein feedback from the robot encodersallows reconstruction of the robot’s path in real-time. Using a joystick interfaced to the computerbase-station, students command the robot remotely over the bluetooth connection while theodometry data is transmitted from the robot to Matlab where the path taken by the robot isplotted in real-time. By driving the robot along the perimeter of the classroom and aroundobstacles, an occupancy-map of the classroom is created. As a final lesson, the so called “parkingproblem” of the differential-drive robot is studied. The students are asked to think of a strategyfor the robot to autonomously drive from a start position and orientation to a goal position andorientation. A controller based on the kinematic model of the robot is implemented anddemonstrated in class. The robot takes the goal position and orientation as input, and usingodometry as feedback, smoothly steers to the desired pose. The performance of the feedbackcontroller is compared to manual control where students use the joystick to control the robot.The above set of in-class lessons, demonstrations, and activities illustrate applications ofmathematical concepts, primarily geometry and trigonometry, to solve practical problems inmobile robotics. Through these lessons, students see practical value in subject areas that mayotherwise appear to be abstract. The full version of the paper will include classroom assessmentof the aforementioned activities and recommendations for future work.