Lab 7: Mobile Robot II

2.12/2.120 Intro to Robotics
Spring 2026¹

Table of Contents

• 1 Mobile Robot
  • 1.1 Understand robot/
  • 1.2 Understand Odometry
  • 1.3 U-Turn
  • 1.4 Circle
• 2 Joystick Control
  • 2.1 Get Robot MAC Address
  • 2.2 Get Controller MAC Address
  • 2.3 Validate Controller
  • 2.4 Run Controller
  • 2.5 Run Joystick Control
• 3 Custom Trajectory
• X Optional Extensions
  • X.1 IMU
  • X.2 Mecanum Wheels

We have already written most of the code for this lab. We hope that you will use the extra free time to fully understand the mobile robot codebase and prepare for the final competition. This lab should take no more than 1 hour to complete. Have a great spring break!

1 Mobile Robot

Estimated time: ≤25 minutes

1.1 Understand navigation code in robot/

Take some time to understand robot_main.cpp, robot_drive.cpp, robot_motion_control.cpp, and robot_wireless.cpp. At a high level:

• robot_main.cpp: Includes the setup() and loop() functions, telling the microcontroller exactly what to do and when.
• robot_drive.cpp: Sets up the motors and implements a PI controller to follow velocity setpoints.
• robot_motion_control.cpp: Calculates odometry and setpoints based on either joystick or a given trajectory.
• robot_wireless.cpp: Sets up two-way wireless communication with and sends messages to the microcontroller on your controller. The messages are defined in lib/wireless.h.

1.2 Understand Odometry

Open robot_motion_control.cpp and read through updateOdometry(). Make sure that you understand exactly how this function calculates odometry data. For reference:

1.3 U-Turn

Make sure your PlatformIO environment is set to be env:robot and upload the code to the microcontroller on your mobile robot. Once the code has finished uploading, unplug the robot from your computer, set it on the ground, power it on, and press RST. Your robot should now autonomously follow a (counter-clockwise) U-Turn!

Your U-Turn will probably not be perfect! While odometry is straightforward to implement, it suffers from problems such as position drift due to wheel slippage. IMU data (or some combination of IMU and odometry) will likely be more reliable.

1.4 Circle

In robot_motion_control.cpp, comment out #define UTURN and uncomment #define CIRCLE. This will change the followTrajectory() function to follow a circle instead of a U-Turn. Your robot should now autonomously follow a circle!

✅ CHECKOFF 1 ✅
Tell us how the code works, specifically how the robot odometry is implemented using the Jacobian. Why is the U-turn inaccurate, and how could we improve its accuracy?

2 Joystick Control

Estimated time: 20 minutes

2.1 Get Robot MAC Address

In order to establish wireless communication, we first have to make sure that both microcontrollers know each other's MAC addresses.

Run lib/Wireless/examples/get_mac.cpp (you will have to temporarily move the existing files inside the src/robot/ folder somewhere else, like the src/temp folder, and replace them with get_mac.cpp). Open lib/Wireless/wireless.h and change robotAddr to the MAC address being printed to the Serial monitor.

If you don't see anything printing, make sure you have selected the right microcontroller port (it must be connected to your laptop via USB). Switch back to "Auto" when you're done with the Serial monitor.

2.2 Get Joystick Controller MAC Address

Connect to the microcontroller on your joystick controller board and change your PlatformIO environment to be env:controller.

Forget how to change environments?

Please refer to the instructions from Lab 6.

Run lib/Wireless/examples/get_mac.cpp (you will have to temporarily move the existing files inside the src/controller/ folder and replace them with get_mac.cpp). Open lib/Wireless/wireless.h and change controllerAddr to this MAC address.

2.3 Validate Joystick Controller

Run src/test_controller/controller_test.cpp. You should see joystick readings in the range [-1.0, 1.0] being printed to your Serial monitor.

2.4 Run Joystick Controller

Upload controller_main.cpp and controller_wireless.cpp to the microcontroller on your controller. This will read the joystick and set up two-way wireless communication with the microcontroller on the mobile robot.

2.5 Drive the Mobile Robot

In robot_motion_control.cpp, comment out #define CIRCLE and uncomment #define JOYSTICK. This will change the followTrajectory() function to follow a joystick instead of a circle.

Set your PlatformIO environment back to env:robot. Upload robot_main.cpp, robot_drive.cpp, robot_motion_control.cpp, and robot_wireless.cpp to the microcontroller on your mobile robot. Make sure the joystick microcontroller is receiving power from a laptop or USB-C power adapter. At this point, you should be able to drive your mobile robot around with your joystick!

✅ CHECKOFF 2 ✅
Show your mobile robot in action to a TA or LA.

3 Custom Trajectory

Estimated time: 15 minutes

In robot_motion_control.cpp, comment out #define JOYSTICK and uncomment #define YOUR_TRAJECTORY. In the followTrajectory() function, make your own path using a state machine, taking UTURN as inspiration. You do not need to make it too complicated :)

✅ CHECKOFF 3 ✅
Show your mobile robot in action to a TA or LA.

X Optional Extension (eXtra Credit)

X.1 IMU

Install an IMU on the circuit board (if not already present), and use the IMU yaw angle to increase the accuracy of robotMessage.theta in the updateOdometry() function of robot_motion_control.cpp. For example, you could make theta completely dependent on the IMU readings. This will probably be very useful for your final project, especially if you use dead reckoning, which is extremely sensitive to deviations in theta.

Potential extension for final project: Mecanum Wheels

Install 2 motors on the front of the robot, and replace all 4 existing wheels with mecanum wheels! Mecanum wheels allow the robot to move in any direction. However, the odometry and controller will be slightly different.

Footnotes

1. Version 1 - 2016: Peter Yu, Ryan Fish and Kamal Youcef-Toumi
   Version 2 - 2017: Yingnan Cui and Kamal Youcef-Toumi
   Version 3 - 2019: Jerry Ng
   Version 4 - 2023: Joseph Ntaimo, Kentaro Barhydt, Ravi Tejwani, Kamal Youcef-Toumi and Harrison Chin
   Version 5 - 2024: Jinger Chong, Josh Sohn
   Version 6 - 2025: Roberto Bolli Jr., Kaleb Blake
   Version 7 - 2026: Stephan Stansfield ↩