From 204161f1ff8829f8f93a552e597ea08767ddda84 Mon Sep 17 00:00:00 2001 From: Steve Drew Date: Tue, 3 Mar 2026 22:40:45 -0700 Subject: [PATCH 1/3] Enrich README with background, motivation, findings, and preprint badge (closes #10) - Add Research Square preprint badge alongside CI badge - Expand intro paragraph to highlight EDSim's high-fidelity testbed value - Add Background & Motivation section covering ED crowding problem, limits of traditional simulation, and LLM-driven cognitive architecture - Add Key Findings & Validation section with three evidence-backed bullets (wait-time distributions, authentic behavior, rapid experimentation) - Update Table of Contents to include the two new sections Co-Authored-By: Claude Sonnet 4.6 --- README.md | 47 ++++++++++++++++++++++++++++++++++------------- 1 file changed, 34 insertions(+), 13 deletions(-) diff --git a/README.md b/README.md index 6e51a88..d97f632 100644 --- a/README.md +++ b/README.md @@ -1,23 +1,44 @@ # EDSim: An LLM-Powered Emergency Department Simulation Using Generative Agents [![CI](https://github.com/denoslab/EDSim/actions/workflows/ci.yml/badge.svg)](https://github.com/denoslab/EDSim/actions/workflows/ci.yml) +[![Preprint](https://img.shields.io/badge/preprint-Research%20Square-blue)](https://www.researchsquare.com/article/rs-8960989/v1) -EDSim is a multi-agent simulation of emergency department (ED) workflows driven by large language model (LLM)-powered autonomous agents. Each agent (doctors, nurses, patients) perceives its environment, makes decisions through cognitive modules, and interacts with other agents in real time, producing realistic ED dynamics suitable for operational analysis and research. +EDSim is a multi-agent simulation of emergency department (ED) workflows driven by large language model (LLM)-powered autonomous agents. Each agent — doctor, bedside nurse, triage nurse, or patient — perceives its environment, makes decisions through cognitive modules, holds natural-language conversations with other agents, and executes clinically-constrained behaviors in real time. The result is a high-fidelity testbed for ED operations research that goes beyond what traditional simulation methods can produce. + +## Background & Motivation + +Emergency departments face chronic crowding and complex patient-flow challenges that are difficult to study safely in real clinical settings. Traditional simulation approaches — discrete-event models and rule-based agent simulations — can reproduce coarse statistics like average wait times and throughput, but they cannot capture the fine-grained staff behaviors, spontaneous communication, and dynamic decision-making that ultimately shape patient outcomes. + +EDSim addresses this gap by grounding each virtual agent in an LLM-driven cognitive architecture adapted from the [Generative Agents](https://arxiv.org/abs/2304.03442) framework (Park et al., 2023). Agents form memories, reflect on past events, plan upcoming actions, and converse with one another — all while operating under clinically realistic constraints such as CTAS triage protocols and role-specific responsibilities. This combination enables the simulator to surface emergent workflow patterns that simpler models miss. + +The immediate practical value is speed and safety: hospital managers and researchers can run what-if experiments — reallocating beds, adjusting staffing levels, changing triage thresholds — in minutes on commodity hardware, without disrupting real patient care. + +## Key Findings & Validation + +EDSim has been validated against real-world ED operational data (details in the [preprint](https://www.researchsquare.com/article/rs-8960989/v1)): + +- **Realistic wait-time distributions** — Baseline simulation results align with historical patient wait-time distributions when stratified by CTAS triage acuity level (1–5). +- **Authentic agent behavior** — Agents generate convincing clinical conversations and exhibit plausible adaptive behaviors under novel workflow conditions not seen during prompt design. +- **Rapid experimentation** — Operational interventions (e.g., bed reallocation, surge staffing) can be evaluated end-to-end in minutes, enabling iterative hypothesis testing. + +> EDSim represents a new paradigm for healthcare operations research — combining data-driven modeling with LLM-generated behavior to produce a simulator that is both statistically grounded and behaviorally rich. ## Table of Contents -1. [Quickstart](#quickstart) -2. [Architecture Overview](#architecture-overview) -3. [Agent Roles](#agent-roles) -4. [Installation](#installation) -5. [Running the Simulation](#running-the-simulation) -6. [Configuration](#configuration) -7. [Data Collection](#data-collection) -8. [Testing](#testing) -9. [CI/CD](#cicd) -10. [Acknowledgments](#acknowledgments) -11. [Citation](#citation) -12. [License](#license) +1. [Background & Motivation](#background--motivation) +2. [Key Findings & Validation](#key-findings--validation) +3. [Quickstart](#quickstart) +4. [Architecture Overview](#architecture-overview) +5. [Agent Roles](#agent-roles) +6. [Installation](#installation) +7. [Running the Simulation](#running-the-simulation) +8. [Configuration](#configuration) +9. [Data Collection](#data-collection) +10. [Testing](#testing) +11. [CI/CD](#cicd) +12. [Acknowledgments](#acknowledgments) +13. [Citation](#citation) +14. [License](#license) ## Quickstart From 29f030abd21346049c97a520c6a2b5f94f25af48 Mon Sep 17 00:00:00 2001 From: Steve Drew Date: Tue, 3 Mar 2026 22:43:55 -0700 Subject: [PATCH 2/3] Update README title and remove Generative Agents references - Change heading to match paper title: "An Agentic Simulator for Emergency Department Operations" - Remove inline Generative Agents framework link from Background section - Rephrase Architecture section to avoid framework reference - Update Acknowledgments to reflect clinical contributors Co-Authored-By: Claude Sonnet 4.6 --- README.md | 10 ++++------ 1 file changed, 4 insertions(+), 6 deletions(-) diff --git a/README.md b/README.md index d97f632..a76b342 100644 --- a/README.md +++ b/README.md @@ -1,4 +1,4 @@ -# EDSim: An LLM-Powered Emergency Department Simulation Using Generative Agents +# EDSim: An Agentic Simulator for Emergency Department Operations [![CI](https://github.com/denoslab/EDSim/actions/workflows/ci.yml/badge.svg)](https://github.com/denoslab/EDSim/actions/workflows/ci.yml) [![Preprint](https://img.shields.io/badge/preprint-Research%20Square-blue)](https://www.researchsquare.com/article/rs-8960989/v1) @@ -9,7 +9,7 @@ EDSim is a multi-agent simulation of emergency department (ED) workflows driven Emergency departments face chronic crowding and complex patient-flow challenges that are difficult to study safely in real clinical settings. Traditional simulation approaches — discrete-event models and rule-based agent simulations — can reproduce coarse statistics like average wait times and throughput, but they cannot capture the fine-grained staff behaviors, spontaneous communication, and dynamic decision-making that ultimately shape patient outcomes. -EDSim addresses this gap by grounding each virtual agent in an LLM-driven cognitive architecture adapted from the [Generative Agents](https://arxiv.org/abs/2304.03442) framework (Park et al., 2023). Agents form memories, reflect on past events, plan upcoming actions, and converse with one another — all while operating under clinically realistic constraints such as CTAS triage protocols and role-specific responsibilities. This combination enables the simulator to surface emergent workflow patterns that simpler models miss. +EDSim addresses this gap by grounding each virtual agent in an LLM-driven cognitive architecture. Agents form memories, reflect on past events, plan upcoming actions, and converse with one another — all while operating under clinically realistic constraints such as CTAS triage protocols and role-specific responsibilities. This combination enables the simulator to surface emergent workflow patterns that simpler models miss. The immediate practical value is speed and safety: hospital managers and researchers can run what-if experiments — reallocating beds, adjusting staffing levels, changing triage thresholds — in minutes on commodity hardware, without disrupting real patient care. @@ -65,7 +65,7 @@ EDSim consists of three main components: ### Backend Simulation Engine (`reverie/`) -The core engine that drives agent behavior. Each agent is equipped with cognitive modules that mirror the generative agents architecture: +The core engine that drives agent behavior. Each agent is equipped with cognitive modules that form the simulation's cognitive loop: - **Plan** — generates and revises daily and immediate action plans - **Perceive** — observes nearby agents, objects, and events in the ED environment @@ -320,9 +320,7 @@ git tag v1.0.0 && git push origin v1.0.0 ## Acknowledgments -EDSim builds on the **Generative Agents** framework by Park et al. (Stanford/Google): - -> Joon Sung Park, Joseph C. O'Brien, Carrie J. Cai, Meredith Ringel Morris, Percy Liang, and Michael S. Bernstein. 2023. *Generative Agents: Interactive Simulacra of Human Behavior.* In Proceedings of the 36th Annual ACM Symposium on User Interface Software and Technology (UIST '23). ACM. +We thank the clinical staff and operational teams who contributed domain expertise to the design of EDSim's agent behaviors and workflows. ## Citation From 1e29e7866672459efc0836c6e860470835b467d0 Mon Sep 17 00:00:00 2001 From: Steve Drew Date: Tue, 3 Mar 2026 22:47:24 -0700 Subject: [PATCH 3/3] Add DOI, license, Python, Docker, and journal badges to README Co-Authored-By: Claude Sonnet 4.6 --- README.md | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/README.md b/README.md index a76b342..9b35c28 100644 --- a/README.md +++ b/README.md @@ -1,7 +1,12 @@ # EDSim: An Agentic Simulator for Emergency Department Operations [![CI](https://github.com/denoslab/EDSim/actions/workflows/ci.yml/badge.svg)](https://github.com/denoslab/EDSim/actions/workflows/ci.yml) +[![Docker](https://github.com/denoslab/EDSim/actions/workflows/docker.yml/badge.svg)](https://github.com/denoslab/EDSim/actions/workflows/docker.yml) [![Preprint](https://img.shields.io/badge/preprint-Research%20Square-blue)](https://www.researchsquare.com/article/rs-8960989/v1) +[![DOI](https://img.shields.io/badge/DOI-10.21203%2Frs.3.rs--8960989%2Fv1-blue)](https://doi.org/10.21203/rs.3.rs-8960989/v1) +[![npj Digital Medicine](https://img.shields.io/badge/npj%20Digital%20Medicine-under%20review-orange)](https://www.nature.com/npjdigitalmed/) +[![Python 3.9](https://img.shields.io/badge/python-3.9-blue.svg)](https://www.python.org/downloads/) +[![License](https://img.shields.io/badge/License-Apache%202.0-blue.svg)](LICENSE) EDSim is a multi-agent simulation of emergency department (ED) workflows driven by large language model (LLM)-powered autonomous agents. Each agent — doctor, bedside nurse, triage nurse, or patient — perceives its environment, makes decisions through cognitive modules, holds natural-language conversations with other agents, and executes clinically-constrained behaviors in real time. The result is a high-fidelity testbed for ED operations research that goes beyond what traditional simulation methods can produce.