Project Goals and Vision

This project, "AI for Thai EMS," is a research initiative by the National Institute for Emergency Medicine (NIEM) to create a framework for applying Artificial Intelligence (AI) to enhance Emergency Medical Service (EMS) processes in Thailand. The core vision is to address persistent challenges in the current EMS system, such as complex data management, delays in emergency response, and errors in data processing for service reimbursement.

The project aims to leverage AI to improve the efficiency and accuracy of emergency incident assessment and command. By automating and assisting in data extraction and analysis, the system intends to reduce human error, speed up response times, and create a more transparent process for both incident management and billing. The application serves as a showcase for this framework, demonstrating the practical application of AI in a simulated emergency dispatch scenario.

Sources: App.tsx:23-33, metadata.json:2-5, components/Header.tsx:8-9

Core Project Objectives

The project is guided by four primary objectives that define its scope and intended outcomes.

Sources: constants.ts:3-20, App.tsx:35-41

AI in Action: The Vision

The project's vision is demonstrated through two key features: an "AI in Action" emergency dispatch simulation and an "EMS AI Assistant" chatbot. These components illustrate how AI can be integrated into the EMS workflow.

Sources: App.tsx:43, App.tsx:98

Emergency Dispatch Simulation

The AiDemo component provides a real-time simulation of an AI-assisted emergency call. It visualizes a conversation between a dispatcher and a caller, while an AI model analyzes the transcript as the call progresses.

The simulation flow is as follows:

1. The user starts the simulation.
2. The application displays a pre-defined conversation transcript line by line.
3. After each line from the "Caller," the accumulated transcript is sent to the Gemini AI for analysis.
4. The AI returns structured data, which is then displayed on the "AI Analysis Dashboard."
5. The process repeats until the conversation ends.

Sources: components/AiDemo.tsx:72-85, components/AiDemo.tsx:132-134

This diagram illustrates the simulation and analysis loop within the AiDemo component.

AI Analysis and Data Structuring

The core of the AI's function is to transform unstructured conversation text into structured, actionable data. The analyzeEmergencyCallWithGemini function sends the transcript to the Gemini API with a specific JSON schema, forcing the model to extract key information.

The expected output is defined by the AiAnalysisResult interface.

// From types.ts
export interface AiAnalysisResult {
  incidentType: string | null;
  location: string | null;
  severity: 'Critical' | 'Urgent' | 'Non-Urgent' | null;
  recommendedAction: string | null;
  extractedData: {
    name: string;
    age: string;
    condition: string;
  }
}

Sources: types.ts:11-21

The following diagram shows the data model for the AI's analysis output.

The sequence diagram below details the interaction between the frontend, the service layer, and the Gemini API during the analysis process.

EMS AI Assistant Chatbot

The Chatbot component provides another view of the project's vision: an interactive AI assistant. Users can ask questions in text or via audio, and the AI responds. This demonstrates a more conversational and on-demand application of AI for EMS personnel or the public.

The chatbot features several modes to enhance its capabilities:

• Thinking Mode: Uses the more powerful gemini-2.5-pro model for complex queries.
• Search: Integrates Google Search to ground responses with real-world data.
• Maps: Integrates Google Maps for location-based queries.

Sources: components/Chatbot.tsx:10, components/Chatbot.tsx:199-201, services/geminiService.ts:85-87

Project Implementation Plan

The project is structured into three main activities, each contributing a certain percentage to the overall completion.

Sources: constants.ts:31-47, components/Timeline.tsx

The following diagram shows the sequential flow of the project's main activities.

Expected Impacts

The project anticipates delivering significant benefits across public, private, and community sectors.

Sources: constants.ts:50-69, App.tsx:63-95

Conclusion

The "AI for Thai EMS" project outlines a clear vision for modernizing Thailand's emergency medical services through the strategic application of artificial intelligence. By focusing on tangible objectives—developing an AI system, improving data workflows, creating a governance framework, and reducing errors—the project aims to build a faster, more accurate, and more reliable EMS system. The included demonstrations of an AI-assisted dispatch and a conversational AI assistant provide a concrete look at how this vision can be realized, promising significant positive impacts for government agencies, private industry, and the public.

Sources: App.tsx:23-33, constants.ts:3-20, constants.ts:50-69