Traditional approaches for test cell automation are inefficient and insufficient for today's fast paced, ever-changing testing requirements. Future test cell environments will utilize software plug-ins that are easily accessible and configurable by test engineers; they will also include automated test systems that provide for easy integration and quick development of these plug-ins. Consequently, software plug-ins need to be a primary consideration during design of test programs rather than a secondary feature. Such a design approach can greatly increase flexibility for test cell automation as well as post processing while maintaining data integrity. Southwest Research Institute has developed a data acquisition and processing system incorporating software plug-ins to automate our test cell operations. This paper highlights our approach for software plug-in handling. Current test cell automation systems usually rely on hardware plug-in modules for physical I/O. In most cases, these can be easily swapped out and added to by the test engineer. To accommodate the rate at which today's testing environments change, automated test systems must also widely offer software plug-in capabilities. These plug-ins lower upfront software development costs, through improved software reuse, while also providing greater long term adaptability. One popular example of software plug-in use is Hardware-In-the-Loop models deployed in a test cell automation system. To avoid needing a software developer for every change of a given model, test engineers must have the capability to load and swap models at run time. Software plug-ins are also important for adding custom user interfaces, device communication logic, custom control logic and custom calculations. Especially with custom calculations, software plug-ins provide a way to maintain data reproducibility and data integrity while offering a heightened level of flexibility. This aspect relies on the automated test system having a robust strategy for plug-in handling, archival, and retrieval as well as a well-defined connection to data post processing. Automated test systems and data post processor utilities can be made to share software plug-ins, thus adding to data reproducibility benefits. Calculation validation efforts are reduced as the calculations only need to be vetted once rather than once for each execution environment. The main benefit of plug-ins is to bring software modularity, a basic concept for software developers, to the forefront of deployed applications and thereby dramatically increasing the flexibility possible for each deployment scenario.
