As you all know, robots can be controlled mainly in the following modes:

• Manual: In manual control, the robot is controlling manually by a human. The controlling is done using a remote controller or teach pendant.
• Semiautonomous: The semiautonomous robot will have both manual and autonomous control. For simple task, it can work autonomously but in complex task it may change its mode to manual.
• Fully autonomous: An autonomous robot has complete control over its action and can think for itself. It can learn and adapt, and very much everything is controlled by the robot itself.

We can choose the model of robot control based on our application. In this chapter, we are mainly discussing implementing a manual robot control; we can call it distance control or teleoperation. In teleoperation, the robot and human can be far apart, and the operator may not able to see the real robot moving but may get some visual feedback. Rather than manual control, some teleoperated robots have different levels of autonomy integrated. The robots can take action entirely by following the commands sent by the operator or only receiving high-level commands and taking care of other stuff autonomously.

We are going to be discussing a project to teleoperate a robot using hand gestures. The major component that we are using to detect the gestures is an inertial measurement unit (IMU). The IMU is fitted into a hand glove, and with specific hand gestures, we can move or rotate the robot. The project uses the Arduino-ROS combination to compute IMU orientation and send it to PC. A ROS node runs on the PC, which maps the orientation data into twist messages (geometry_msgs/Twist), which is the command velocity of the robot. We will look at more analysis of the project design in the upcoming sections.

We are going to discuss following topics in this chapter:

• Teleoperating a TurtleBot using a keyboard
• Gesture teleop: teleoperating using hand gestures
• Setting up the project
• Interfacing the IMU MPU-9250 with the Arduino and ROS
• Visualizing the IMU TF on Rviz
• Converting IMU data into twist messages
• Integration and final run
• Teleoperating using an Android phone