Establishing an Efficient Test Framework for Embedded Software Verification via Hardware-in-the-Loop Testing 2012-01-2029

The heavy-duty diesel industry continues to expand the use of sophisticated electronic controls to reduce greenhouse gas emissions and improve fuel economy in the transportation sector. This inevitably leads to a need for increased knowhow in all aspects of embedded software development, e.g. in writing model-based specifications, rapid controls prototyping, automatic code generation, and hardware-in-the-loop (HiL) testing. In addition, the software development organization has to recognize the need to establish maturity in its verification and validation test framework, which includes, among other things, a proper definition of testing process and workflow, greater organizational focus, and a robust tool architecture for software testing.

Navistar has already undertaken a model-based approach to software development and continues to improve their processes, technology, and tooling in this area. Another focus area in the past year has been HiL-based test framework for increased software testing in the lab environment. In this paper, we will outline a HiL-based verification and validation (V&V) framework-developed jointly by Navistar and ETAS-that is built upon open-architecture tools, standard interfaces, and established industry best-practices. The primary goal of the HiL-based test framework was to create a foundation for a mature test process based on an environment where software testing is tightly coupled with software requirements. Further, we will demonstrate how Navistar, through the adoption of this test framework, is now able to fully automate functional testing of embedded control unit (ECU) software, re-use test cases across ECU families and test platforms, improve regression test coverage, and automatically test the diagnostics software in an engine ECU. Thanks to this framework, Navistar software engineering teams in diverse geographies are now able to increase collaboration, resulting in improved software quality with lower overall V&V test effort.