Intern
    Data Science Chair

    ERP Fraud Game

    A financial game with opportunities for fraud

    As part of our efforts to detect fraudulent behavior in Enterprise Research Planning (ERP) system data in the research project DeepScan, we designed a game for the generation of innovative fraud cases in cooperation with the Games Engineering lecture chair of the university of Würzburg.

    Inspired by the serious game ERPSim, the three-player online multiplayer game lets players take managerial positions in a cereal production company.

    While players are encouraged to cooperatively run the company by buying raw materials, producing cereals and offering their products on a simulated market, the game's design also aims to give players the degrees of freedom to deviate from normal business processes and illicitly obtain company assets for their personal enrichment.

    The general game design idea is described in detail in our contribution to the IEEE Conference on Games 2021 with the title "A financial game with opportunities for fraud".
    Further, we design a first prototype that models the normal business process, and show that data generated from this prototype mimics realistic ERP system data and may prove useful in the application of machine learning based fraud detection as part of the DeepScan project.

    The first Unity-based prototype that includes normal business behavior and a first, explicitly modeled type of fraud, can be played in a 3-player multiplayer setting or locally with AI collaborators.
    The game is publicly available and can be downloaded under the following link:

    For further questions and feedback regarding the game, feel free to contact us under: erp-fraud-game@informatik.uni-wuerzburg.de