SafePath

Objective:

This project focuses on developing a data product that aids users in navigating London safely by optimizing routes based on real-time and historical crime data. The system integrates descriptive, predictive, and prescriptive analytics to provide actionable insights for route safety, emphasizing user interactivity, efficiency, and scalability.

Key Design Features:

1. Interactive Map-Based Interface:

   • Users can input start and end locations, visualizing multiple routes overlaid on a map styled with CartoDB Positron tiles.

   • A heatmap highlights high-crime areas for situational awareness.

2. Safety Scoring:

   • Routes are ranked based on a safety score that incorporates crime proximity, recency, and severity.

   • Spatial operations using sf ensure efficient geospatial processing.

3. Real-Time Data Integration:

   • The system fetches route data dynamically via the OpenRouteService API.

   • Crime data is processed monthly, ensuring up-to-date insights.

4. Dynamic Visualizations:

   • Heatmaps, route polylines, and interactive tables present data intuitively.

   • Color-coded routes and pop-ups provide detailed route summaries, including distance, duration, and safety scores.

5. Scalability:

   • Modular design allows adaptation to other cities or datasets with minimal changes.

   • Efficient spatial joins and buffered operations ensure performance with large datasets.

Design Process:

1. Requirement Analysis:

   • Identified user needs: safety prioritization, route visualization, and ease of use.

   • Defined goals for interactivity and efficiency.

2. Data Acquisition and Preprocessing:

   • Crime datasets from multiple months were merged and cleaned, retaining relevant columns (Month, Latitude, Longitude, Crime type).

   • Recency weights were computed for predictive analytics.

3. Back-End Development:

   • Implemented geospatial operations using sf for proximity analysis and risk scoring.

   • API integration with OpenRouteService to fetch routing data and provide alternative paths.

4. Front-End Development:

   • Designed an intuitive Shiny interface with responsive elements.

   • Interactive features include search inputs, dynamic tables, and map overlays (heatmaps, routes).

5. Testing and Validation:

   • Validated accuracy of safety scores and ensured robustness against edge cases (e.g., missing data).

   • Performance optimized for real-time updates with large datasets.

6. Deployment:

   • The modular structure ensures easy deployment and adaptability to other regions or environments.