Team Cereberus

What it does

Find My Force is a system that helps detect and estimate the location of unknown RF signals. It collects signal snapshots from receivers and processes them through a machine learning pipeline to classify signals and estimate their origin.

The platform includes:

A signal ingestion service that receives IQ snapshots from receivers
A machine learning service that analyzes signal characteristics
A geolocation service that estimates the probable location of the transmitter
A frontend map interface that visualizes signals and estimated positions

Together, these components allow users to monitor incoming signals and see where they may be originating from.

How we built it

We built Find My Force using a microservices architecture.

Core components:

Frontend: A web interface with an interactive map built using MapLibre to visualize signal detections.
Ingest Service: Receives and stores IQ snapshot data from sensors.
ML Service: Processes the signal data to classify or extract features from RF snapshots.
Geolocation Service: Uses signal properties and receiver metadata to estimate the transmitter location.
Aggregation Service: Combines outputs from the ML and geolocation services and delivers results to the frontend.

The services communicate through APIs, allowing the system to remain modular and scalable.

Challenges we ran into

One of the biggest challenges was dealing with raw RF signal data and turning it into something useful for downstream services. IQ data is noisy and complex, which made feature extraction and classification difficult.

Another challenge was combining outputs from different services in a meaningful way. Synchronizing signal snapshots, ML predictions, and geolocation estimates required careful data handling.

We also had to ensure that the system remained responsive enough for real-time visualization, which meant optimizing the pipeline and keeping services lightweight.

Accomplishments that we're proud of

We’re proud that we were able to design and implement a working end-to-end pipeline during the hackathon.

Specifically:

Building a modular microservice architecture
Successfully ingesting and processing real signal snapshots
Visualizing results on an interactive map
Integrating machine learning with geolocation analysis

Seeing signals appear on the map with estimated origins was a big milestone for the team.

What we learned

This project taught us a lot about RF signal processing, especially how challenging it is to work directly with IQ data.

We also gained experience designing scalable microservice systems, coordinating multiple services, and building data pipelines that can process streaming data.

Another key takeaway was the importance of clear interfaces between services, which allowed different parts of the system to evolve independently.