Mobility / crowdsourced route safety AIurban mobility and cycling safetyParticipation documented

CycleAI

Lisbon, Portugal

CycleAI states that it collaborates directly with users through crowdsources and uses the perception poll and hotspot map to train and validate its safety model.

human-centered designcivic data contributiondata collectiontraining data generationvalidationdeployment

Project documentation

Overview

Region: Europe
Lead organization: CycleAI
Organization type: nonprofit / social enterprise
Technology group: route safety prediction and recommender
Activity status: active
Start year: 2020
Last updated: Mar 25, 2026

Participation documentation

Participation mode: crowdsourced safety perception and hotspot reporting
Participants: cyclists and micromobility users
Methods: crowdsourcing; perception polls; hotspot mapping; user-contributed data
AI lifecycle stages: data collection; training data generation; validation; deployment
Evidence summary: CycleAI states that it collaborates directly with users through crowdsources and uses the perception poll and hotspot map to train and validate its safety model.
Atlas assessment: Cautious · Medium confidence · evidence grade A
Uncertainty: Participation is clear but mostly contributory rather than shared decision-making.