ball-balancing 3-RRS (Revolute-Revolute-Spherical) Platform. It can dynamically balance a ball in real time and perform motion tasks such as moving in straight lines, quadrants, and even circles. The robot uses real-time computer vision, inverse kinematics, and PID control for its operation.
- Python
- SolidWorks
- 3d Printing
- OpenCV (Contour Detection, HSV Tuning)
- Inverse Kinematics (for 3-RRS)
- PID Control
- Arduino Board
| S.No | Component | Quantity | Purpose |
|---|---|---|---|
| 1. | Arduino R4 Minima | 1 | Microcontroller |
| 2. | Power Supply Or 5V batteries | 1 | Servo Control |
| 3. | **5MP Webcam Camera ** | 1 | Vision (ball detection) |
| 4. | Usb Cable | 1 | Connects Laptop to Arduino |
| 5. | RKI1206 Servo Motors | 3 | Actuation of platform |
| 6. | Ball Joint (M6) | 3 | Platform joints (RRS design) |
| 7. | Ball Bearings (ID - 6mm, OD - 16mm) | 6 | For joint rotation |
🧮 Inverse Kinematics (Snippets)
We derived and implemented the IK for the 3-RRS platform to calculate the exact servo angles required to achieve desired platform tilt and height. This allowed precise control over the ball's motion.
- HSV Trackbar for color tuning
- Faster response than BGR and no noise Like Grayscale ,for robust tracking
cv2.findContoursandcv2.minEnclosingCircleto detect the ball- Real-time position and area returned for control logic
after manual tuning and testing, a finely tuned PID controller was implemented, keeping the ball at centre.
📐 CAD Files (Link)
All the 3D models used in this project—including the base, platform, servo holders, and linkages—were custom-designed using Fusion360. You can find the .stl files in the STLs directory. These files are ready for 3d printing.