Translator, Editor: Sophie

Source: http://makezine.com

As an Arduino and robotics enthusiast, have you always dreamed of creating a cool little car?A father made an upgraded off-road balancing car to meet his son’s outdoor wishes, which can autonomously navigate, adapt to various uneven terrains, and can even climb!

Basic Project Introduction

Project Name: High-Precision Autonomous Navigation Balancing Car

Production Time: 3-8 hours

Production Cost: $400-$500

Difficulty Level: Difficult

ArduRoller is a self-balancing, anti-fall robot that can quickly identify obstacles and navigate autonomously. It surpasses traditional cars, usable in multi-directional recognition, uneven terrains, and sloped scenarios.

The core of the self-balancing car is the Inertial Measurement Unit (IMU), which contains three single-axis accelerometers and three single-axis gyroscopes. The accelerometers detect the acceleration signals of the object, while the gyroscopes detect the angular velocity signals of the carrier relative to the navigation coordinate system, measuring the object’s angular velocity and acceleration in three-dimensional space, calculating the object’s posture. These nine sensors sample at a frequency higher than 1000 times per second, combined with a mathematical filter called DCM, which can obtain data such as angles and rotation rates through DCM for balancing.

To enable navigation, it is essential to know the car’s position and its destination accurately. However, GPS can only be precise to a few meters, while the car needs higher precision. Wheel encoders allow centimeter-level accuracy and complement GPS well.

The author chose the 3D Robotics APM 2.5 autopilot because it integrates all necessary sensors, all software is free and open-source, and it is convenient to make. The ArduPilot system developed by the DIY drone community can quickly complete the autonomous navigation car robot.

Image and Text Tutorial Area

1. Material Collection: Motors and Wheels

Choose brushed DC motors with encoder units; a 34-1 gear reducer can effectively increase torque. Additionally, off-road wheels can help the car adapt to various uneven terrains.

2. 3D Print the Chassis Parts

The 3D model files of the chassis parts can be downloaded from the Thingiverse website.

3. Assemble the Motors

Install the motors into the 3D printed base, using hot melt glue to prevent the motors from loosening.

4. Circuit Connection

As shown in the diagram, the electronic components include APM open-source autopilot, motors, logic level converters, Arduino mini boards, etc.; connect the GPS module and remote control radio module for autonomous operation; to achieve manual control of the car, an R/C receiver needs to be added, and then connect the jumper of the APM module with the PPM module.

(Note: Some components in the circuit diagram come from SparkFun, users can choose similar hardware replacements as needed.)

Fix the electronic components on the foam board; connect the mainboard to the motors; place the entire foam board in the middle of the chassis; ensure that the “front” arrow of the autopilot is upright.

5. Assemble the Chassis

The entire chassis is divided into three parts, and they should be assembled together; the battery is hidden under the top round arch decoration; a switch can be added to the side of the car; install the off-road wheels; the LED inside the chassis can be seen through the acrylic glass when lit; install the remote control radio module, which can be used to receive remote tasks or directly control the car.

6. Programming (Click Read Original Text to see the complete download path)

Download the source files (APM 2.5 or 2.6) and upload them to the APM module; use modified Arduino IDE— ArduPilot-Arduino (Windows or Mac);

Download the wheel encoder source files (Arduino Pro Mini) and then upload them to the Arduino Pro Mini board using the normal Arduino IDE;

If you want to achieve autonomous tasks, you can use Mission Planner (PC version, Windows only) or DroidPlanner 2 (Android version).

7. Usage

Once the entire car is completed, it can be controlled using a wireless remote control or given an autonomous task via GPS.

The Mission Planner software can help you track the car’s path, speed, and direction using Google Maps, run your own Python files, download and analyze task logs, etc.

You can add a video transmission module to the car, or use a GoPro to record high-definition videos, even add sonar so that the car can fully implement obstacle avoidance!

About the Author

Jason Short is the design director of 3D Robotics, with over 20 years of product design experience, having worked on consumer product design and user experience research for companies like HTC, Samsung, Sony, and LG. In 2009, he assisted in the development of ArduPilot, an open-source autopilot for R/C aircraft. In 2010, he invented the ArduCopter, an unmanned helicopter that has been widely used worldwide.

*Feel free to share in your circle of friends. For reprinting, please contact for authorization.