A System Design Framework for Digital Process Innovation

点击翻译按钮获取中文摘要

Many of the processes (e.g. commerce, education, healthcare) in modern society are moving from physical to digital execution. Digitization, broadly defined as the conversion of analog information and physical processes into digital format, is foundational to a digital society. On one hand, digitization can enable new connectivity across all components of a digital ecosystem (e.g. agents, processes, technologies) to enhance functionality beyond what is enabled in physical process execution. We refer to this process enhancement as digital process innovation (DPI). However, DPI projects will have varied complexity with respect to the degree of connectivity across different ecosystem components. Our inability to account for this complexity leads to hazards and other unintended consequences from DPI. To support information system (IS) design for DPI we need to first characterize a digital ecosystem according to the agents, processes and coordination that exists within the ecosystem and then identify IS design requirements for DPI. However, DPI can be designed in many ways creating a multi-agent decision problem. To address that issue, we need to design flexible tools to enable IS designers to simulate potential IS design configurations to establish common ground about IS design requirements. The long-term goal of this proposed research program is to develop a system design framework for DPI. Three specific objectives will accomplish the overall goal. The first objective is to model and characterize a DPI ecosystem using ontological engineering. Reasoning capabilities will be developed to characterize an ecosystem as a tight or loose coupled system based on the ontology inputs. The second objective will operationalize the ontology into a multi-agent system (MAS) architecture. The dynamic nature of a DPI ecosystem means the MAS architecture can be configured in different ways and IS designers must establish common ground on the architecture for a specific DPI project. The third objective will implement the MAS architecture as a computer based multi-agent decision support system (MADSS). MADSS will enable IS designers to simulate different variations of the MAS architecture to support the development of common ground on IS design requirements for a DPI project. Proof of concept usability testing of the MADSS will be done with case studies of tight coupled (online education) and loose coupled (community-based healthcare planning) domains. The result of this research will be an innovative framework to support DPI. It will have a theoretical impact by developing a method to model and characterize a DPI ecosystem as tight or loose coupled. It will have a practical impact by providing a flexible tool that enables IS designers to simulate different DPI configurations to help them to establish common ground about IS design requirements to support digital process innovation.