That’s It?

The idea is to create a self balancing robot, which should learn to balance itself without explicitly coding in the logic to do so. In a traditional sense we need to apply reinforcement learning model to let the robot train itself. Instead of going into the details of how that can be accomplished, here I lay down the foundations of the hardware setup required to achieve this.

General Hardware Overview

On a high level we have the following:

• Micro-controller - Orchestrate input from different sensors and drive the motors appropriately.
• Gyroscope - Sensor informing about the tilt angle of the robot.
• Stepper Motor - Precisely control the movement of the robot.

The diagram below shows a schematic connection of how different components can be connected together: Little Step(per)