A robot developed by [Nikodem Bartnik] demonstrates that machine learning techniques can be applied to lighter-weight hardware. The robot autonomously navigates a simple racetrack using a two-wheeled design with tank-style steering. The Arduino Uno controls the robot, which utilizes a Slamtec RPLIDAR sensor to map out its surroundings. The microcontroller is equipped with a Bluetooth link and an SD card for data storage.
To train the machine learning model, the robot was manually driven around the racetrack multiple times while collecting LIDAR data. Control inputs were recorded along with the data to create a dataset. Feature selection techniques were then implemented to refine the collected data points to those most relevant for completing the driving task.
[Nikodem] explains the process of creating and refining the machine learning model to enable the robot to drive autonomously on various racetrack designs. The experiment serves as a primer on the application of machine learning techniques to small embedded platforms.
Machine learning is a field of artificial intelligence that focuses on the development of algorithms and models capable of learning from and making predictions or decisions based on data. It involves the processing of large datasets to identify patterns and make accurate predictions or take appropriate actions.
The Arduino Uno is a popular microcontroller board that allows for the control of electronic components and devices. It is widely used in various projects for its ease of use and versatility.
The Slamtec RPLIDAR sensor is a laser scanner commonly used in robotics applications for mapping and navigation purposes. It provides real-time 360-degree scanning and accurate distance measurement capabilities.