An interactive R Shiny web application that visualizes crime patterns across London using real-time data from official UK police databases. Built as the capstone project for DATA*6500, the dashboard combines spatial analysis, data visualization, and user-centered design to provide actionable insights for residents, policymakers, researchers, and law enforcement.
Live App: renmotc.shinyapps.io/newhorizon/
- Project Background
- Executive Summary
- Technical Implementation
- Key Insights & Findings
- Technical Specifications
- Team Collaboration & Development Process
- Future Enhancements
- Sources
This dashboard was developed during the summer semester of the DATA*6500 program. Our team designed an interactive Shiny app that ingests, processes, and visualizes 100,000+ crime records across multiple categories, delivering real-time insights into London's crime landscape.
The London Crime Map Dashboard transforms raw crime data into a dynamic visualization platform.
It enables:
- Crime category filtering (e.g., theft, burglary, vehicle crime, public order)
- Temporal analysis with interactive time-series charts
- Spatial insights via heat maps, clustering, and choropleth mapping
- Customizable exploration from borough-level summaries to individual incident details
Target users:
Residents | Policymakers | � Researchers | Law Enforcement
- Ingested from UK Police API + London DataStore
- 100k+ records standardized and cleaned
- Processing pipeline included:
- Geocoding validation
- Temporal normalization
- Statistical aggregation at borough + LSOA levels
- Data quality controls for missing values and inconsistent formats
- Built with R Shiny’s reactive programming model
- Key functionalities:
- Dynamic date filtering
- Multi-category selection
- Geographic boundary selection
- Visualization modes:
- Point mapping
- Choropleth density maps
- Time-series trends
- Responsive UI/UX design
- Leaflet mapping with clustering for performance optimization
- Heat maps + Kernel Density Estimation
- Distance-based analysis from landmarks
- Overlays for administrative regions
- Clean, professional design
- Progressive disclosure: Start broad → drill down into details
- Accessibility-first color palette & hierarchy
- Central boroughs (e.g., Westminster, Camden) show consistently high crime volume
- Suburban areas (e.g., Kingston upon Thames) experience unexpected category spikes
- Seasonal trends:
- Property crimes ↑ in winter
- Public order offenses ↑ in summer
- Democratizes access to complex data for community decision-making
- Languages & Tools:
R,R Shiny,Leaflet,Plotly,DT,dplyr,ggplot2 - Data Sources:
- Deployment: ShinyApps.io with CI/CD pipeline
- Performance Optimizations: caching, lazy loading, clustering
- Security: data anonymization + session management
- Predictive Analytics → ML models to forecast hotspots
- Real-time Data Streaming → live feeds + alerts
- Mobile App → location-based safety alerts
- Socioeconomic Data Integration → richer context for policymakers
- Public API → for researchers & third-party developers