SMMARTS: An Open Architecture Development Platform for Modular, Mixed, and Augmented Reality Procedural and Interventional Simulators.

INTRODUCTION: Different simulators often share elements, resulting in different laboratories doing redundant work. This can lead to higher development and acquisition costs, proprietary, incompatible technology, lack of interoperability, and large inventories that reduce accessibility to the benefits of simulation. Simulation technology can become more affordable and scalable with open architecture and modular design. We describe the System of Modular Mixed and Augmented Reality Tracking Simulators (SMMARTS) open architecture, rapid development platform for designing and building modular procedural and guided-intervention simulators. METHODS: A modular stand provides mechanical indexing (registration) of a modular anatomical block representing the anatomy relevant to the simulated intervention. A software development kit (SDK) integrated with the hardware (stand and hand-held tracked tools such as a needle and ultrasound probe) facilitates software development. The SMMARTS SDK at https://github.com/UF-CSSALT/SMMARTS-SDK developed in Unity Technologies' Unity game engine includes Arduino microcontroller and NDI's 6 degrees of freedom tracking connectivity along with software tools such as a replayer, user interface templates, 3D visualization of the virtual counterparts of physical elements, scoring monitors, cognitive aids, common error messages, and Experience Application Programming Interface compatibility. RESULTS: We used SMMARTS to develop 9 different simulators internally (instructor-less central venous access currently deployed to Iraq, prostate biopsy, epidural loss-of-resistance, ventriculostomy, pterygopalatine fossa block, lumbar/chronic pain blocks, chest tube insertion) and externally (intravenous access). DISCUSSION: As a living tool, SMMARTS now has sufficient functionality and benefits that we can share it to help clinicians and engineers focus more on content specific to learning objectives rather than back-end tasks.