The Pendulum Series
A pendulum looks simple. Hang a rod from a bearing and let it swing. Yet it holds most of what makes robots hard: a model that is never quite right, sensors that never quite tell the truth, and a system that falls over the moment you try to hold it upright.
This series takes one steel rod and a high-resolution encoder all the way from a free swing to balancing upside down.
| Part | Question |
|---|---|
| 1. Model, Sensor and Error | How well does a model derived on paper predict a real swinging rod, and how noisy is the sensor watching it? |
| 2. ROS 2 Integration | How do you get every encoder reading onto a robot’s software stack, on time and without dropping any? |
| 3. Kalman Filter, Hanging | Can blending the model with the sensor give a better estimate of angle and speed than either alone? |
| 4. Controlled Hanging Pendulum | With a motor at the pivot, how do you damp the swing, or pump it higher? |
| 5. Inverted Pendulum | How do you balance the rod at the top, where gravity works against you? |