Barcelona Accessibility Smart System: Machine Learning-Driven Inclusive Urban Planning

This article introduces the Barcelona Accessibility Smart System, an application that uses machine learning technology to predict urban accessibility. Through geospatial feature analysis, the system provides data-driven insights for urban planners and decision-makers, aiming to improve the travel experience of all citizens (especially groups such as people with mobility impairments and the elderly) and promote inclusive urban planning. Core content includes system background, technical methods, application scenarios, and social significance.

Urban accessibility involves multi-dimensional challenges such as physical environment (sidewalks, slopes, etc.), transportation facilities (bus stops, subway elevators), public services (building entrances, restrooms), and information access (braille signs, audio prompts). Traditional assessments rely on manual surveys, which have problems like limited coverage, delayed updates, strong subjectivity, and high costs, making it difficult to meet the needs of real-time city-wide assessments.

The system integrates multiple technologies:

1. Machine learning models: Uses Random Forest (to handle high-dimensional geospatial data, capture non-linear relationships, and provide feature importance analysis) and SMOTE technology (to solve class imbalance issues); 2.Geospatial analysis: Extracts spatial features, models location relationships, calculates accessibility, and identifies problem hotspots; 3.Data visualization: Accessibility heatmaps, path analysis charts, facility distribution maps, and trend analysis; 4.User-friendly interface: Simple operation process, fast analysis, and multi-format export (CSV, GeoJSON, etc.).

System Requirements: Supports Windows10+, macOS10.14+, Ubuntu20.04+; Minimum 4GB RAM (8GB+ recommended); At least 500MB disk space; Depends on Python3.x and libraries like NumPy, Pandas, scikit-learn, Matplotlib. Installation Steps: Visit the GitHub project's Releases page → Download the corresponding version installer → Install as prompted → Launch the app and load data → Run analysis and visualize → Export results. Supported Data Formats: CSV (structured attributes), GeoJSON (geospatial data).

The system provides support for multiple parties:

• Urban planners: Priority ranking (identify areas needing improvement), program evaluation, progress tracking;
• Public service optimization: Transportation planning (optimize bus routes), facility layout, emergency evacuation route planning;
• Citizen services: Accessible travel recommendations, facility queries, problem feedback collection.

The prediction accuracy of the system depends on data quality and model selection. High precision can be achieved when using high-quality datasets, highlighting the core position of data quality in machine learning applications—high-quality algorithms cannot compensate for the defects of poor-quality data.

The system can be expanded in the future:

1. Real-time data integration (connecting to urban sensor networks);
2. Multi-modal analysis (integrating image recognition to evaluate accessibility);
3. Personalized recommendations (customizing routes based on user needs);
4. Crowdsourced data (citizens participating in updating accessibility information).

The Barcelona Accessibility Smart System embodies the concept of technology for good. Improving urban accessibility through machine learning not only helps specific groups but also enhances urban inclusiveness and livability. AI does not replace human judgment; instead, it provides data support to help decision-makers make more informed choices and promote social equity.