Table of Contents
Fetching ...

Using Hybrid System Dynamics and Discrete Event Simulations to Identify High Leverage Targets for Process Improvement in a Skill-based Organizational Structure

Eric Enos, Daniel Herber

TL;DR

The paper tackles delays in delivering work within a skill-based IT organization by introducing a hybrid simulation approach that combines System Dynamics and Discrete Event Simulation. By applying the method to a healthcare provider's IT case study, it reveals how project and operational work interact, creating coupled queues and nontrivial wait times, especially as teams with different skills collaborate. The study identifies leverage points, such as reducing rework, limiting switching between tasks, and increasing automation (CI/CD), to improve throughput and quality in a constrained time horizon. The findings offer practical governance and automation strategies for complex IT environments and lay the groundwork for extending the models to multi-team settings and staffing interventions.

Abstract

This paper is based on a case study of an IT organization in a large, US-based healthcare provider, and develops simluation models to identify areas for performance improvement. These organizations are often grouped into departments by technical skill and support both operational work (tickets) and project work (tasks) of various priorities. From a practical standpoint, resource managers and staff regularly manage all work as queued and assign / complete it based on the priorities of the day. Using project and operational metrics from the case study organization, the hybrid model using both system dynamics and discrete event simulation developed through this research depicts the flow of work through a skill-based team as well as many of the key factors that influence that workflow, both positive and negative. Experience indicates that the interaction between project and operational work -- as well as between teams with differing skills -- entangles work queues and wait times within those queues in a way that rapidly scales in complexity as the number of interacting individuals and teams increases. Results from model simulation bear out this intuition. Scaling the models to accommodate multiple teams is a topic of future research.

Using Hybrid System Dynamics and Discrete Event Simulations to Identify High Leverage Targets for Process Improvement in a Skill-based Organizational Structure

TL;DR

The paper tackles delays in delivering work within a skill-based IT organization by introducing a hybrid simulation approach that combines System Dynamics and Discrete Event Simulation. By applying the method to a healthcare provider's IT case study, it reveals how project and operational work interact, creating coupled queues and nontrivial wait times, especially as teams with different skills collaborate. The study identifies leverage points, such as reducing rework, limiting switching between tasks, and increasing automation (CI/CD), to improve throughput and quality in a constrained time horizon. The findings offer practical governance and automation strategies for complex IT environments and lay the groundwork for extending the models to multi-team settings and staffing interventions.

Abstract

This paper is based on a case study of an IT organization in a large, US-based healthcare provider, and develops simluation models to identify areas for performance improvement. These organizations are often grouped into departments by technical skill and support both operational work (tickets) and project work (tasks) of various priorities. From a practical standpoint, resource managers and staff regularly manage all work as queued and assign / complete it based on the priorities of the day. Using project and operational metrics from the case study organization, the hybrid model using both system dynamics and discrete event simulation developed through this research depicts the flow of work through a skill-based team as well as many of the key factors that influence that workflow, both positive and negative. Experience indicates that the interaction between project and operational work -- as well as between teams with differing skills -- entangles work queues and wait times within those queues in a way that rapidly scales in complexity as the number of interacting individuals and teams increases. Results from model simulation bear out this intuition. Scaling the models to accommodate multiple teams is a topic of future research.
Paper Structure (18 sections, 2 figures)

This paper contains 18 sections, 2 figures.

Figures (2)

  • Figure 1: Vensim system dynamics model of a single team reflecting both project and operational workflows.
  • Figure 2: Simevents queuing model depicting entity generators for each work type feeding four engineers with individual queues.