WDR (Walk. Detect. Report)

Inspiration

As college students, we have all faced moments where we felt unsafe walking home late at night, navigating unfamiliar areas, or simply not knowing which routes were secure. Many times, people only learn about incidents through word of mouth or social media after they have already occurred.

We wanted to change that.

WDR (Walk Detect Report) was inspired by our own experiences living on campus and being placed in uncomfortable or unsafe situations. We realized that awareness is often the first step toward prevention. Our goal was to create a platform that empowers people to stay informed, detect risks early, and feel safer in their environments.

WDR makes safety visible, accessible, and community-driven.

What it does

WDR (Walk Detect Report) is a real-time incident map that allows users to report safety incidents by voice or text, view live heatmaps of nearby reports, direct SMS reports, and stay alert about potential risks in their area.

It promotes awareness and prevention through:

Walk: Navigate safely with a visual understanding of your surroundings, using real, mapped locations across Rutgers–New Brunswick.

Detect: Identify patterns and hotspots of unsafe activity through dynamic heatmaps that update in real time.

Report: Instantly share incidents using quick text input, SMS, or hands-free voice commands. When someone sends a report via SMS, it is automatically geolocated to the exact Rutgers area they’re in and displayed live on the map.

WDR also uses AI through Hugging Face to categorize incidents into assault, harassment, lighting issue, suspicious behavior, or other, ensuring that users can understand and respond appropriately to different types of safety concerns.

By combining these features, WDR turns community awareness into a shared safety network, helping users stay calm, informed, and connected.

How we built it

WDR combines mapping, voice processing, and live reporting into a single real-time platform that visualizes safety information for users.

Frontend

We built the frontend using:

React for responsive and modular component-based design.

Leaflet.js for lightweight, interactive maps.

OpenStreetMap for open-source geographic visualization and location data.

These tools allowed us to render dynamic heatmaps, show user-submitted reports, and display location data in an intuitive and fast interface.

Backend

The backend was developed with:

Node.js and Express to handle routing, data flow, and report management.

Uses JSON test data to synchronize user reports across clients.

Hugging Face Inference API to process voice inputs and extract key information from natural language reports, automatically classifying them into categories such as assault, harassment, lighting issue, suspicious behavior, or other.

Uses Twilio SMS communication APIs for the SMS feature, enabling users to directly text their report from their phone. The system then maps the report’s real location across Rutgers–New Brunswick and displays it instantly on the platform.

Challenges we ran into

Real-time synchronization: Ensuring that user-submitted reports appeared instantly from SMS communication.

Voice recognition accuracy: Background noise and inconsistent phrasing made reliable natural language processing challenging.

Balancing design and emotion: We wanted the UI to communicate urgency without causing panic, emphasizing trust and clarity.

Privacy concerns: Protecting user data while keeping reports location-relevant required careful anonymization strategies.

Accomplishments that we're proud of

Building a fully functional prototype that updates incident reports live on the map.

Successfully integrating voice-based reporting powered by Hugging Face NLP models.

Creating a minimal, calming interface that enhances usability in stressful situations.

Designing a project that addresses a real-world safety issue experienced by students daily.

Mapping real locations and safety incidents across Rutgers–New Brunswick, connecting the community through shared awareness.

What we learned

We learned how to integrate multiple technologies into a seamless safety application, combining mapping, real-time data, and AI-driven voice processing.

Key takeaways:

Designing for accessibility and emotional calmness was a key factor for this project.

Using open-source tools like Hugging Face and OpenStreetMap offers flexibility and scalability for safety-focused projects.

Building around user trust and data ethics is critical for sensitive applications.

What's next

Our next steps include:

Mobile app development for iOS and Android with push notifications for nearby alerts.

Integration with public safety APIs to include verified law enforcement data.

Machine learning enhancements to automatically classify and prioritize incident reports.

Community-driven features such as safe route recommendations and trusted reporter verification.

WDR aims to become a trusted platform for college and city safety, starting with Rutgers–New Brunswick and expanding outward — creating an informed, connected, and protected community.