Rapid Street Assessment Tool

🚧 THIS IS A WORK IN PROGRESS.

Overview

Rapid Street Assessment (RSAs) are designed to provide quick, comprehensive analysis of street and land use data - using USRNs (Unique Street Reference Numbers).

It consists of a Python backend using Robyn framework and a frontend built with Taipy GUI.

Example API Call Flow:

Core Features

1. Street Information Analysis

• Fetches and analyses street network data and special designations
• Provides detailed information about:
  • Street characteristics
  • Special designations and restrictions
  • Engineering difficulties
  • Traffic sensitivity
  • Street Manager Aggregated Stats
• Uses chat gpt-4o mini to generate human-readable analysis of the technical data

2. Land Use Analysis

• Retrieves and processes land use and building data
• Provides insights about:
  • Property types and distributions
  • Land use categories
  • Total area statistics
  • Building characteristics
• Uses chat gpt-4o mini to generate human-readable analysis of the technical data

3. Collaborative Street Works Analysis

• Retrieves and processes street manager data, street information and land use data and combines them into a single object,
• Merges insight from land use and street informatioon to provide a recommendation for collaborative street works
• Uses chat gpt-4o mini to generate human-readable analysis of the technical data

Technical Architecture

Frontend (Taipy GUI)

Currently just a very basic frontend to test the backend - this will be improved.

Backend (Robyn)

• RESTful API endpoints:
  • /street-info and /street-info-llm: Summary of network and RAMI data as well as street manager stats
  • /land-use-info and /land-use-info-llm: Summary of Land use and building information
  • /collaborative-street-works-llm: Collaborative street works recommendation endpoint
• Asynchronous processing of multiple OS NGD API calls
• Intelligent data filtering and data aggregation
• Integration with OpenAI's chat gpt-4o mini for data interpretation

Key Dependencies

• Python ≥3.11
• Robyn (API framework)
• Taipy (GUI framework)
• LangChain (AI processing)
• MotherDuck (data storage)

Data Sources

• Ordnance Survey National Geographic Database (NGD)
• Supports multiple OS data collections:
  • RAMI (Routing and Asset Management Information)
  • Network data
  • Land use data
• Street Manager data from MotherDuck
• Street Impact scores from MotherDuck (created monthly by myself)