Rooftop Optical Link With Adaptive Beam Pointing

ISEF Category: Embedded Systems

Subcategory: Optics  ·  Difficulty: Advanced  ·  Setup: University Lab  ·  Time: Full Year

The Hook

A tiny misalignment can kill a rooftop light link faster than a broken wire. Wind moves buildings, even when you cannot see it. Your job is to keep a beam locked on target while the world keeps shaking it loose. That turns optics, control systems, and embedded code into one real engineering problem.

What Is It?

This project studies free-space optical communication, which sends data through air with light instead of copper or fiber. Think of it like a flashlight conversation between rooftops. If the beam lands on the detector, the link works. If wind sway nudges the transmitter or receiver out of line, the signal drops.

The hard part is beam pointing. A PID controller is a feedback system that keeps correcting error, like a driver making small steering moves to stay in a lane. A learned PID controller adds data-driven tuning, so the system can adjust its response when the mount vibrates, the wind shifts, or the alignment drifts over time. Your project tests whether adaptive control keeps the optical link steadier than fixed control.

Why This Is a Good Topic

This is a strong science fair topic because you can measure it clearly. You can track signal strength, packet loss, pointing error, and recovery time under different sway conditions. The topic also connects to a real engineering need, since optical links matter in last-mile communication, drones, satellites, and secure data transfer. You can learn control loops, sensor feedback, signal processing, and experimental design in one project.

Research Questions

• How does adaptive PID tuning change received signal strength compared with fixed PID control under the same sway pattern? ?
• What is the effect of wind-mimicking motion frequency on link outage time for a rooftop optical terminal? ?
• Does adding a learning step to the controller reduce pointing error after repeated alignment disturbances? ?
• To what extent does detector aperture size change tolerance to beam jitter in a free-space optical link? ?
• Which control strategy, open loop, fixed PID, or learned PID, gives the lowest packet loss during intermittent motion? ?
• How does beam divergence affect the tradeoff between alignment tolerance and received optical power? ?

Basic Materials

• Collimated LED source or laser diode module with safe enclosure.
• APD receiver module or photodiode receiver module.
• FPGA board or Pi Pico.
• Motorized pan-tilt mount or servo gimbal.
• Inertial sensor or accelerometer for motion sensing.
• Breadboard, jumper wires, and connectors.
• Oscilloscope or logic analyzer.
• Laptop for code upload and data logging.
• Tripod or rigid mounting hardware.
• Safety glasses matched to the light source.

Advanced Materials

• Avalanche photodiode receiver with transimpedance amplifier.
• FPGA development board with high-speed I/O.
• Pi Pico or microcontroller for secondary control.
• Precision pan-tilt stage with encoder feedback.
• Beam profiler or optical power meter.
• Quadrant photodiode or position-sensitive detector.
• Wind tunnel access or controllable vibration table.
• High-speed data acquisition system.
• Optical filters matched to the source wavelength.
• Rigid optical breadboard and isolation mounts.

Software & Tools

• Python: Analyzes signal stability, control performance, and error patterns across trials.
• ImageJ: Measures beam spot position and size from aligned test images.
• MATLAB or GNU Octave: Models control loops and compares response curves.
• QGroundControl: Helps with logging and testing hardware-style control interfaces if you prototype motion control.
• KiCad: Helps you design support circuits and connector layouts for the receiver or controller.

Experiment Steps

1. Define the single performance metric you care about most, such as link uptime, pointing error, or received power.
2. Separate the optical path, motion system, and controller so you can test each one without guessing where failures start.
3. Choose a baseline control strategy first, then define what adaptive learning must improve.
4. Plan a repeatable sway challenge that changes one motion variable at a time.
5. Build a measurement plan that records both beam position and communication quality at the same time.
6. Decide how you will compare controller versions with the same alignment test and the same analysis method.

Common Pitfalls

• Using a detector that saturates or clips, which hides real changes in received power.
• Testing only one motion pattern, which makes the controller look better than it is on other sway conditions.
• Letting ambient room light leak into the receiver, which adds noise and masks the optical signal.
• Tuning the controller on the same data you use to judge success, which inflates performance claims.
• Mounting the transmitter or receiver on a flexible support, which makes the platform move more than the beam target.

What Makes This Competitive

A competitive version of this project would compare more than one control method under the same motion stress test. You could report not just average signal strength, but also outage duration, recovery speed, and error after each disturbance. Strong projects also separate hardware limits from control limits, so you can tell whether the weak link is the optics, the sensing, or the algorithm. That kind of analysis looks much deeper than a simple demo.

Project Variations

• Test the same beam-pointing system with visible LEDs versus infrared LEDs to compare alignment tolerance and detector noise.
• Swap rooftop sway for tabletop vibration and compare whether the controller works better on one motion spectrum than the other.
• Keep the optics fixed and compare a simple PID controller with a learned controller across different receiver aperture sizes.

Learn More

• MIT OpenCourseWare: Search for controls engineering, feedback systems, and signal processing lecture notes and assignments.
• NASA Technical Reports Server: Search for papers on optical communication, pointing, acquisition, and tracking systems.
• NOAA National Weather Service: Use local wind data to design realistic sway scenarios for outdoor testing.
• PubMed: Search for review articles on photodetector noise, optical sensing, and embedded control in wearable or field systems.
• IEEE Xplore: Search for papers on free-space optical communication and beam tracking, then filter for review articles and conference papers.