State Machine to SysML v2 State Machine via Graph Recognition

State Machine diagrams are a core tool in a system engineersĺ toolset. These diagrams represent the state of a system over time, as well as the transitions between states due to events or conditions of the overall system. State Machines are often used in the development of vehicle and flight software, and their associated systems, and can be invaluable aids for engineers to reason about the behavior of a system due to varying inputs and constraints.Countless state diagrams have been generated over the past 30 years during the development, acquisition, fielding and upgrading of defense systems. But like many digital engineering artifacts, these diagrams are likely stuck as an image embedded in a PDF, or if delivered as a model, are locked down in a proprietary data format in the modeling tool used by the contractor.SysML v2, with its textual notation based syntax, has the potential to liberate all defense system models from vendor locked tools and formats, and can be viewed and modified in a tool as simple as NotePad. In this proposed effort, we will evaluate and develop an image recognition to generative SysML v2 textual notation pipeline to automate the conversion of state machine diagrams into an open and reusable SysML v2 textual notation format, the future format of Model Based Systems Engineering.The capabilities in this proposed effort will be built on our team's platform SysGit, the first git based SysML v2 modeling software suite. SysGit is purpose built to work with your enterpriseĺs existing git-based tooling, seamlessly integrating into the advanced version control and merging capabilities of GitLab, GitHub and open source tools like Gitea, inheriting your enterprise security postures and quickly integrating into your DevSecOps pipelines.SysGit already includes automated generation of SysML v2, and is purpose built to achieve nearly 100% compliant syntax for SysML v2 requirements from input requirement text. Building on top of these base capabilities in SysGit, our team will first develop an image recognition and parsing pipeline that will ingest an image of a state machine, and extract information on the structure and content of the input state machine. This extracted information will then be fed into our existing generative SysML v2 AI pipeline, Plato, outputting a textual notation model of the state machine diagram.