CycloSense is a concept project that explores how machine learning and IoT data can inform safer urban mobility.

The design system bridges hardware-level data and user-facing insight, transforming complex motion datasets into clear visual narratives for both individual cyclists and city infrastructure teams.

The Challenge

Lisbon faces a growing number of bike thefts, and more can be done to mitigate and prevent them.
Additionally, the solution could also help analyze biker behavior and improve urban mobility data.

Main issues identified:

1. Rising bike thefts

2. No reliable tracking

3. No usage insights

Our Approach

CycloSense captures a bike’s motion and orientation data using a gyroscope Arduino + IMU sensor (MPU-6050), which collects angular velocity and acceleration.

This data enables real-time monitoring of bike activity and helps identify unusual movement patterns that may signal theft or irregular use.

Solution: Data Visualisation Dashboard

The CycloSense dashboard displays motion states through a data visualisation interface built with p5.js.
It detects and classifies movement into three distinct states:

1. Stationary → No movement detected

2. In Motion → Normal usage pattern

3. Anomaly → Irregular movement detected

Using motion thresholds and pattern recognition, the system detects potential theft events.

Outcome

CycloSense demonstrates how IoT sensors, machine learning, and UX design can merge to improve urban mobility safety.
It provides a foundation for preventing bike theft and optimising route analytics, supporting both individual cyclists and city planners.

🔗 Live demo: p5.js prototype