Taxi-based Emergency Medical System

TL;DR

This paper proposes a novel, low-cost emergency medical response system utilizing radio-based taxis and reinforcement learning to reduce response times in low-resource settings.

Contribution

It introduces a reinforcement learning framework for deploying taxis as emergency responders, addressing the gap in timely medical aid in low and middle income countries.

Findings

Demonstrates feasibility of taxi-based emergency response system

Shows potential for reduced response times

Provides a reinforcement learning approach for optimal dispatching

Abstract

In case of a severe accident, the key to saving lives is the time between the incident and when the victim receives treatment from the first-responders. In areas with well designed emergency medical systems, the time for an ambulance to arrive at the accident location is often not too long. However, in many low and middle income countries, it usually takes much longer for an ambulance to arrive at the accident location due to lack of proper services. On the other hand, with ubiquitous wireless connectivity, and emergence of radio based taxis, it seems feasible to build a low-cost emergency response system based on taxi service. In this report, we explore one such solution for deployment of a taxi-based emergency response systems using reinforcement learning.