A hybrid process-mining approach for simulation modeling

Waleed Abohamad, Ahmed Ramy, Amr Arisha

    Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

    Abstract

    This paper presents a hybrid Modeling and Simulation framework to address business process challenges. The framework has integrated process mining techniques in the conceptual modeling phase to support developing simulation models that are unbiased and close reflection of reality in a timely manner. The hybrid approach overcomes the pitfalls of traditional conceptual modeling by using process mining techniques to discover valuable process knowledge from the analysis of event logs. The proposed hybrid framework has been applied to an Emergency Department (ED) in order to identify performance bottlenecks and explore improvement strategies in an attempt to meet national performance targets. A large number of unique process flows (i.e. patient pathways) within the ED were uncovered and deviations from process guidelines were accurately identified. Results show that unblocking of ED outflows have a significant impact on patients length of stay (over 80% improvement) rather than increasing the ED physical capacity.

    Original languageEnglish
    Title of host publication2017 Winter Simulation Conference, WSC 2017
    EditorsVictor Chan
    PublisherInstitute of Electrical and Electronics Engineers Inc.
    Pages1527-1538
    Number of pages12
    ISBN (Electronic)9781538634288
    DOIs
    Publication statusPublished - 28 Jun 2017
    Event2017 Winter Simulation Conference, WSC 2017 - Las Vegas, United States
    Duration: 3 Dec 20176 Dec 2017

    Publication series

    NameProceedings - Winter Simulation Conference
    ISSN (Print)0891-7736

    Conference

    Conference2017 Winter Simulation Conference, WSC 2017
    Country/TerritoryUnited States
    CityLas Vegas
    Period3/12/176/12/17

    Fingerprint

    Dive into the research topics of 'A hybrid process-mining approach for simulation modeling'. Together they form a unique fingerprint.

    Cite this