Emergency Services Deployment optimisation using DSM for maximizing coverage, minimizing the response time for Fatal Accidents on Hyderabad Outer Ring Road 🛣️

Refer to the Execution Guide PDF/Doc for using the modules

This repo contains:

• Algorithms and Source Codes
• Outputs as .csv
• Execution Guide
• Reference materials
• Presentation discussing the solution

EMS Fleet Optimization Quick Guide - EfkonStrabag

Welcome to the Ambulance Optimization repository! This project aims to optimize the deployment of ambulances on highways using accident clustering data and mathematical optimization.

Overview

This repository contains several modules to process raw data, cluster demand points, calculate travel times, and optimize ambulance deployment using the PuLP library. Below is a step-by-step guide on how to run each module on a custom dataset.

Table of Contents

• Pre-processing
• Clustering
• OD Time Matrix
• PuLP Optimization Module
• Additional Resources
• Installation
• Contributing
• License

---

1. Pre-processing

Description

This module maps accident chainage to the nearest equipment's latitude and longitude. Use this step if your dataset lacks accident lat-lon and frequency.

Input Fields

• Accident_Chainage
• Accident_Frequency
• Equipment_Chainage
• Equipment_Lat
• Equipment_Lon

Output

• .csv file with Accident_Chainage, Accident_Lat, Accident_Lon

Execution Steps

1. Prepare your raw dataset with the mandatory fields.
2. Run the pre-processing script:

   python pre_processing.py --input <path_to_raw_dataset> --output <path_to_processed_output>

2. Clustering

Description

This module clusters the processed accident data to form demand points.

Input Fields

• Accident_Chainage
• Accident_Lat
• Accident_Lon

Output

• .csv file with Cluster_ID, Demand_Point_Lat, Demand_Point_Lon, Demand_Frequency

Execution Steps

1. Ensure you have the processed dataset from the Pre-processing step.
2. Run the clustering script:

   python clustering.py --input <path_to_processed_dataset> --output <path_to_demand_points_output>

3. OD Time Matrix

Description

This module calculates the shortest travel time from each ambulance location to each demand point.

Input Fields

• Ambulance locations with fields: Ambulance_ID, Ambulance_Lat, Ambulance_Lon
• Demand points with fields: Demand_Point_ID, Demand_Point_Lat, Demand_Point_Lon

Output

• .csv file with Ambulance_ID, Demand_Point_ID, Travel_Time

Execution Steps

1. Prepare your dataset with ambulance locations and demand points.
2. Run the OD Time Matrix script:

   python od_time_matrix.py --ambulance <path_to_ambulance_locations> --demand <path_to_demand_points> --output <path_to_od_matrix_output>

4. PuLP Optimization Module

Description

This module uses mathematical optimization to maximize the coverage of each demand point with the given primary (r1) and secondary (r2) response times and alpha reliability.

Input Fields

• OD Time Matrix from the previous step with fields: Ambulance_ID, Demand_Point_ID, Travel_Time
• Additional parameters: Primary response time (r1), Secondary response time (r2), Alpha reliability

Output

• Optimized ambulance deployment plan including fields: Ambulance_ID, Assigned_Demand_Point_ID

Execution Steps

1. Ensure you have the OD Time Matrix from the previous step.
2. Define your primary and secondary response times, and alpha reliability.
3. Run the optimization script:

   python pulp_optimization.py --input <path_to_od_matrix> --r1 <primary_response_time> --r2 <secondary_response_time> --alpha <alpha_reliability> --output <path_to_optimization_output>

Additional Resources

• Comparison of each optimization iteration with old ambulance deployment:
  • optimization-comparison.ipynb
• Loop iteration for multiple parameters and visualization:
  • optimisation_pipeline.ipynb

Installation

Clone the repository and install the required dependencies:

git clone https://github.com/AGAMPANDEYY/Ambulance-Optimisation-EfkonStrabag.git
cd Ambulance-Optimisation-EfkonStrabag
source venv-efkon/bin/activate