Logo

Inspiration

Communication is the backbone of emergency response, but traditional systems such as radio can fail because of equipment damage, interference, or harsh conditions.

We wanted to explore whether a simple optical signal—a rapidly flashing light—could act as a reliable backup communication channel.

LumaLink was inspired by one idea:

When radio communication fails, light can still carry a message.

Because missing even a few flashes can change the decoded message, we needed a system with predictable timing. That is where QNX RTOS became essential.

What it does

LumaLink receives high-speed Morse code light flashes using a Raspberry Pi Camera Module 3 Wide connected to a Raspberry Pi 5 running QNX RTOS.

The system:

Detects rapid light flashes
Converts them into Morse code
Decodes the Morse into readable English
Uses a local AI model to interpret the message
Launches an emergency dashboard when an SOS signal is detected
Plays an audible alarm and reads the alert aloud

The full pipeline runs locally without relying on the cloud.

How we built it

LumaLink combines embedded vision, real-time processing, local AI, and a web interface.

We used:

QNX RTOS on a Raspberry Pi 5
Raspberry Pi Camera Module 3 Wide
C with the QNX Camera API
A deterministic Morse decoder
A local Qwen language model
A lightweight Python web server
A browser-based emergency dashboard

The camera application measures the duration of each light pulse:

Short pulse → dot
Long pulse → dash
Longer gap → letter or word separation

The decoded message is then passed to the local AI, which produces a concise emergency interpretation for the operator.

QNX provides predictable scheduling for the camera-processing pipeline, helping reduce timing jitter and missed flashes during rapid optical communication.

Challenges we ran into

The biggest challenge was developing on QNX within a limited hackathon timeframe.

Some of the main difficulties were:

Setting up the QNX development environment
Learning the QNX Camera API
Configuring the Raspberry Pi camera
Detecting short flashes reliably
Distinguishing the flashing source from other light in the environment
Tuning dot, dash, and gap timing
Running a local AI model on embedded hardware
Connecting the camera, decoder, AI, and web server into one pipeline

We also had to make sure the AI did not interfere with the time-critical camera processing. To solve this, the camera pipeline completes first, saves the Morse output, and only then launches the AI.

Accomplishments that we're proud of

Built an end-to-end optical communication system on QNX
Detected computer-generated Morse flashes using a camera
Converted light pulses into readable messages
Ran a local AI model directly on the Raspberry Pi
Created an automatic SOS web dashboard
Added audible and spoken emergency alerts
Built a fully offline system with no cloud dependency
Connected the complete workflow into a single launch command

What we learned

We learned why real-time operating systems matter in vision-based communication.

In a normal operating system, unpredictable scheduling delays can cause dropped or delayed frames. In LumaLink, that could cause a dot to be interpreted as a dash or a letter to be missed entirely.

QNX gives the camera-processing task predictable access to system resources, making the timing pipeline more reliable.

We also learned about:

Embedded computer vision
Real-time scheduling
Camera APIs
Morse timing and signal processing
Local AI inference
Embedded web servers
Designing a complete safety-focused system under time pressure

What's next for LumaLink

Future improvements include:

Better performance in bright or noisy environments
Automatic tracking of moving light sources
Adaptive thresholds for different lighting conditions
Faster optical transmitters
Encrypted optical communication
Support for multiple transmitters
More advanced AI emergency interpretation
Mobile notifications for rescue operators
Integration with maritime and disaster-response systems

Our goal is to develop LumaLink into a reliable backup communication system for situations where radio, cellular, or internet communication is unavailable.