Data Collection in Robotics
Robotics and automation have become increasingly relevant areas of interest in modern business and society. These industries are always in need of more knowledgeable engineers and programmers in order to continue development of new technologies.
FIRST is an organization that seeks to inspire students of all ages to learn about robotics, science, engineering, and technology. FIRST provides many programs which help students learn about these topics and develop real world skills. JumpMind is proud to support these programs not only by sponsoring multiple FIRST teams, but also by providing dedicated mentors. JumpMind has sponsored two teams for the 2018 FIRST Robotics Competition, one of FIRST’s programs directed at high school students. Each year, FRC provides a brand new challenge for which teams have 6 weeks to build and program a robot from the ground up. This year’s challenge is particularly complex and exciting.
Even though the FIRST Robotics Competition is a high school competition, the challenges often require the use of complex programming concepts. Typically, teams will equip their robots with a multitude of sensors to collect real-time data about the robot’s systems and environment. These sensors provide feedback for various control systems and assist with automated processes. For example, teams often use wheel encoders to measure the distance each wheel of the robot has traveled in order to accurately drive specified distances. Many teams use more advanced control systems which might include vision processing based on camera feedback or artificial intelligence. Automated control systems are particularly important in the first 15 seconds of each FRC match, when robots are not allowed be controlled by their human operators.
An important aspect of every control system is the ability to record and analyze data from the inputs and outputs of the system. Many FRC teams develop strategies for collecting and plotting data in order to visualize their control systems. This data can be used to fine tune control systems and improve reliability. Data collection is also crucial for error detection and debugging robot code.
One common use for data collection and analysis in FRC is for a control algorithm called a PID controller. PID (Proportional, Integral, Derivative) controllers are extremely common throughout robotics and automation. A simple example of a PID controller would be cruise control in a modern car. Cruise control allows the driver to set a desired speed, and the PID controller will increase or decrease the motor output based on the cars current speed. The motor output is scaled proportionally by the difference between the desired speed and the current speed, called error. The motor output can also be scaled by the sum of error over time (i.e. integral of error) and the rate of change in error (i.e. derivative of error).
The techniques used by FRC teams for data collection and control systems are not far off from what would be expected in the real world industry. This is just one of the reasons why the programs provided by FIRST are so valuable.
