Facilitation of Designing and Embedding Complex Tests for Real-time Health Management

Abstract

Algorithms and analytic designs for embedded real-time health management are generally implemented in customized manners that are specific to target applications. Such practice largely hinders reuse of analytic designs or parts therein. It also causes significant difficulties for the end users to employ their own algorithms or analytic designs for health management purposes. Qualtech Systems Inc. has developed a tool for implementing embeddable advanced analytic algorithms (tests designs) and executing them in conjunction with the real-time reasoning engine of TEAMS for Real-time and Prognostic Health Management. This new tool, called TEAMS-RTX (TEAMS Real-time, Extended), provides a graphical environment where a user can implement analytic algorithms using mathematical operators and a range of data manipulators (e.g., buffers, concentrators, demultiplexers, etc) and define connectivity with observation streams from monitoring schemes. For facilitating analytic design implementation, a library of mathematical operators is included with TEAMS-RTX. The users can develop their own operators as plug-ins for TEAMS-RTX (algorithms, packaged as C dlls conforming to RTX API) and use those for implementation of analytic designs. TEAMS-RTX tool converts an analytic design into source code (ANSI C, ISO 9899-1990) which is subsequently compiled along with some TEAMS components to result a real-time executable client for a chosen target platform (Windows, UNIX, VxWorks, etc). Using monitored data, this client can provide decisions and information for Real-time and Prognostic Health Management. An easy to customize data ingestion front-end is also developed for allowing the end-user to obtain data directly from sensor interfaces, databases, or files having supported format. This paper provides a detailed account of capabilities, features, interfacing mechanisms, and usability of TEAMS-RTX in real-world health management applications.