Skip to content

Latest commit

 

History

History
62 lines (45 loc) · 1.86 KB

README.md

File metadata and controls

62 lines (45 loc) · 1.86 KB

MobilityAgent

MobilityAgent is an AI-powered transportation analysis system built on crewAI that leverages MobilityGPT for trajectory generation and integrates real-time traffic data for comprehensive mobility solutions in San Francisco.

Architecture Overview

MobilityAgent employs a multi-agent system architecture where specialized AI agents work together to:

  1. Convert natural language locations to road segments
  2. Generate synthetic trajectories using MobilityGPT
  3. Integrate real-time traffic data
  4. Assess route quality
  5. Compile comprehensive reports

Agent Roles & Responsibilities

  • Location Translation Specialist

    • Converts natural language addresses to road segment IDs
    • Uses geospatial mapping and urban network knowledge
    • Ensures accurate location interpretation

  • Trajectory Generation Expert

    • Leverages MobilityGPT for synthetic trajectory creation
    • Simulates realistic traffic patterns
    • Generates route variations based on historical data

  • Traffic Data Specialist

    • Integrates Google Maps API real-time data
    • Monitors current traffic conditions
    • Enriches synthetic trajectories with live data

  • Route Quality Evaluator

    • Assesses generated routes against real-world conditions
    • Analyzes metrics like distance accuracy and time efficiency
    • Provides quality scores for trajectories

  • Transportation Data Integration Specialist

    • Compiles analyses into actionable reports
    • Generates optimization suggestions
    • Creates markdown-formatted comprehensive summaries

Prerequisites

  • Python 3.9+
  • Google Maps API key
  • MobilityGPT model

Installation

  1. Create and activate a new virtual environment:
python -m venv .venv
source .venv/bin/activate 

pip install -e .

Usage

python src/mobilityagent/main.py run