CycloSense
EE 327 Spring 2026
This repository contains the files for the firmware, PCB design, and 3D models used to create our bike computer named CycloSense. CycloSense was created to be a bike computer for the everyday bike user, affordable, easy to use, and includes predictive maintenance alerts and offline navigation.
Objective:
Create a bike computer to provide local navigation, predict maintenence, and display metrics
Steps Taken:
- Load nodes and edges from OpenStreetMap to perform A* algorithm on to find optimal route between two points.
- Convert node and edge data to a 2D grid based on offset from the map origin to save memory and shorten computing time.
- Process audio from brake and classify as squeaky or normal with a ML model trained on Edge Impulse to alert user of necessary maintenance.
- Display speed, time, brake status, and directions overlayed on a map.
- Mount main system to handlebar and microphone + hall sensor to front fork with custom 3D printed parts.
Outcome:
A bike computer prototype that successfully tracks speed and provides navigation while showcasing predictive maintenance using edge AI.
PCB Schematic
PCB Rendering

Full system mounted to handlebar
|

Model of 3D printed enclosure
|