This is what it's all about.
1.5 meters of distance. On dry asphalt, that might feel okay. But in wet conditions, poor visibility, uneven roads, and slippery markings? That same distance becomes life-threatening. The problem: Current sensors are «blind» to their surroundings. They only measure centimeters, but not the actual risk.
We are changing that. As part of my PhD research at the University of Fribourg, I am developing FuzzyCycleSense – the first open-source project that measures cycling safety contextually. We are building a smart sensor box (based on Raspberry Pi & LiDAR) that doesn’t just measure distance, but understands the situation:
Sensors analyze moisture, road conditions, temperature, etc., to calculate the road’s slip risk, for example.
And the human factor? Using a trigger on the handlebars, you mark in real-time whenever you feel unsafe – providing our «Ground Truth.»
Our Goal: We want to scientifically prove that rigid «traffic rules» or recommendations often fail in physical reality. We are turning your bike into a mobile research lab that uncovers «invisible» danger spots in your city. Support me in funding the hardware for our pilot fleet. For evidence-based cycling research at the Uni Fribourg that is grounded in reality and makes our bike lanes truly safer.
What makes my project special
Science over profit: This project is not a private DIY hobby, but the core of my academic research. It’s about gaining independent knowledge for everyone, not creating a commercial product.
Bridging the gap between law and physics: Most traffic concepts rely on rigid rules (e.g., «min. 1.5 meters distance»). But physics is dynamic. Our project brings Fuzzy Logic to cycling for the first time. We calculate risk probabilities based on grip (hygrometry/temperature), vibration, etc.
The human as a sensor (Human-in-the-Loop): Algorithms are great, but your gut feeling is better. Previous trackers ignore your subjective experience. CycleSense gives you a voice: Using the «Safe/Unsafe» button, you validate the data. This allows us to prove for the first time: «The distance was technically maintained here, but due to the wet road, the cyclist felt massively threatened.»
Open Source & Privacy First: No cloud, no subscriptions, no hidden data harvesting. All data is processed locally on the Raspberry Pi Zero (Edge Computing). Blueprints and code (Python) are published entirely to keep our research transparent and reproducible.
This is what I need backing for.
Hardware for reliable research is expensive. I have already developed the prototype (Proof of Concept) and written the software. Now, the goal is to turn the lab setup into a reliable pilot fleet for the streets.
Your contribution goes directly towards:
Hardware components (approx. 50%): LiDAR sensors for millimeter-precise distance measurement, Raspberry Pi as the «brain,» BME280 & IMU sensors for environmental analysis, as well as weatherproof casing and power banks.
The field study (approx. 30%): To make scientifically valid conclusions for my PhD, I need data from many different bicycles. We are equipping a test group in Bern and providing incentives for the participants.
Open-Source development (approx. 20%): Server costs for anonymized data hosting and creating tutorials so other researchers or laypeople can build their own sensors.
Important note: Since this project is running on Science Booster, your support has a massive leverage effect. Together, we are showing that independent safety research matters to the public!