Inspiration

City planning is slow, abstract, and often disconnected from the people who use the streets every day. Dangerous intersections in Toronto and cities worldwide continue to put cyclists and pedestrians at risk, yet decision-makers lack tools to visualize the impact of safety improvements instantly.

Cision was inspired by a personal experience: my cousin Aarav bikes to school near Queen and Spadina. One day, a car rolled through a right turn and missed him by just a meter while he was in the crosswalk. Later, I discovered that intersection was already flagged as high-risk in city data. That moment made the problem personal and clear: data alone is not enough. Cities need a way to see danger and act before accidents happen.

What it does

Cision is an interactive 3D map that instantly redesigns dangerous intersections to be safer. Users can click on high-risk locations and see AI-generated improvements in real time, including:

Clearer signage and lane markings
Protected bike lanes and optimized pedestrian crossings
Multi-stakeholder insights (cyclist, civil engineer, policymaker perspectives)
Actionable safety audits with concrete recommendations

By turning abstract traffic data into a visual, human-centered experience, Cision helps cities prioritize and implement safety measures faster and more effectively.

How we built it

Cision is powered by:

Frontend Framework & UI

Next.js 16
TypeScript
Tailwind CSS 4

Mapping & Visualization

Mapbox GL
React Map GL

AI & Machine Learning

Google Gemini 2.5 Flash
Google Gemini Nano Banana (intersection redesign image generation)
ElevenLabs (real-time voice synthesis for persona agents)
Vercel AI SDK (unified AI integration framework)

Backend & Data

Next.js API Routes (serverless endpoints)
MongoDB (collision data storage and cluster persistence)
Sharp (image processing and panoramic stitching)

Our tech stack combines AI image generation, geospatial analysis, and interactive front-end design to make data instantly actionable for planners, engineers, and policymakers.

Challenges we ran into

Integrating real-world collision data with AI-generated visuals in a way that felt accurate and readable
Designing multi-persona perspectives that were informative yet concise for each type of stakeholder
Ensuring real-time visual updates without performance issues in a web environment

Accomplishments that we're proud of

Created a visual-first urban safety tool that transforms raw data into actionable insights
Successfully implemented AI-powered intersection redesigns that are clear, realistic, and contextually accurate
Built a framework for multi-perspective AI insights, enabling planners to see problems from the eyes of cyclists, engineers, and policymakers
Developed actionable safety audits that go beyond visualization and provide concrete steps for real-world implementation

What we learned

How to merge AI image generation with geospatial and traffic data to create actionable urban planning tools
The importance of visual communication in influencing safety-focused decisions
Balancing real-world accuracy with AI-generated visualizations to build a tool that is both compelling and practical

What's next for Cision

Expand coverage to additional cities and intersections worldwide
Allow users to simulate and compare multiple redesigns for the same location
Integrate predictive modeling to forecast potential accident reductions from proposed changes
Partner with municipal governments and advocacy organizations to deploy Cision as a decision-making tool for real-world urban safety improvements

#UrbanSafety #CivicTech #SmartCities #TrafficSafety #AI #3DVisualization #DataDriven #UrbanPlanning #PedestrianSafety #CyclistSafety #AIForGood #TechForCities #InteractiveMap #CityPlanning #Innovation