Virtualized Embedded Systems for Testing and Development
Reference number | |
Coordinator | Scania CV Aktiebolag - Avd REVE |
Funding from Vinnova | SEK 3 811 236 |
Project duration | March 2014 - August 2017 |
Status | Completed |
Venture | Electronics, software and communication - FFI |
Call | 2013-01925-en |
End-of-project report | 2013-05608eng.pdf (pdf, 1621 kB) |
Important results from the project
The VIRTUES project has been conducted over 3.5 years between Scania CV and KTH to investigate hypotheses about virtualised hardware emulation and testing. The project has developed new software engineering tools that support machine-learning assisted test case generation, as well as virtualised hardware emulation. These tools have been integrated together. The resulting platform has been successfully applied to requirements and fault injection testing of automotive ECU applications.
Expected long term effects
An instruction set simulation platform was constructed and used. A simulation platform for a subset of the truck ECUs was constructed. We have prototyped an integration of the environment model of the HIL rig with the simulation platform, and experimented with novel ways to control the simulation. We have integrated a machine-learning assisted test case generator with the simulator to conduct testing case studies based on different Scania ECU applications. These included requirements and fault injection testing. We have optimised the test case generator for multi-core platforms.
Approach and implementation
The advantages of the VIRTUES approach have been demonstrated, such as lowered costs, flexibility and agility of approach, reliable and rigorous test results, and higher test coverage. Furthermore, we have shown that our techniques scale up to significant problem sizes, although further work is possible to extend scalability even further. By building on the execution platform produced in this project, we can see many interesting next steps. It may also be possible to combine this future work with other related Scania research projects such as ReVamp and Virtual Truck and Bus.