Moving Toward the Obsolescence of Obsolescence: A Walk in the Clouds

Abstract

Many software methodologies advanced by cloud computing can be applied to automatic test software to mitigate test system obsolescence challenges. Classical monolithic test, measurement, and automation systems have traditional challenges when test system components become out-of-date. Distributed computing technologies provide incremental and modular updates that can be applied proactively or reactively to handle equipment failures, shifts in hardware (HW) and software (SW) dependencies, and improved HW and SW components. Advances in cloud computing have driven software technologies that can unlock scaling across hardware systems, introduce service architectures, open new performance possibilities, and decouple historically interlocked components. This paper will address aerospace and defense software challenges, introduce cloud computing, both service-oriented architecture (SOA) and cloud-native microservices, explain underlying principles and tenets, and craft a practical path forward. The cloud will be brought down to earth by showing how some underlying principles and tenets can be used today. These include the single responsibility principle (SRP), DevOps, and gRPC, “a modern open-source high-performance Remote Procedure Call (RPC) framework that can run in any environment.” [1] These practices enable architectures that make obsolescence issues smaller and more approachable. For instance, test hardware and software evolve as available hardware, and the development team's skillsets change over the lifetime of the tester and Unit Under Test (UUT). Using SRP allows individual engineers to troubleshoot isolated portions of complex systems without understanding the entire technology stack. Additionally, partitioning measurement tasks into measurement and analysis subtasks can maximize CPU power by offloading analysis from point-of-use test stations to higher-performance computers. This enables cost savings by removing number crunching from expensive testers and leveraging optimized high-performance computing for analysis tasks. We also show how gRPC natively incorporates low latency, security, and cross-platform interoperability between services. While it is impossible to prevent software, parts, and skills from becoming obsolete, it is possible to mitigate the risks to production through planning, procurement, and system design, including system software design. By walking in the footsteps of cloud computing, we can apply various techniques and technologies from the cloud to automated systems.