Search arXiv⌕ Search
Home/Archive/arXiv

arXiv · 1802.06841

Full Virtualization of Renault's Engine Management Software and Application to System Development

Dirk Von Wissel·Yohan Jordan·Dirk Von·Wissel Renault·Adrian Dolha·Jakob Mauss
View original source ↗Read source document ↗Export metadata ↓

Abstract

Virtualization allows the simulation of automotive ECUs on a Windows PC executing in closed-loop with a vehicle simulation model. This approach enables to move certain development tasks from road or test rigs and HiL (Hardware in the loop) to PCs, where they can often be performed faster and cheaper. Renault has recently established such a virtualization process for powertrain control software based on Simulink models. If the number of runnables exceeds a threshold (about 1500) the execution of the virtual ECU is no longer straight forward and specific techniques are required. This paper describes the motivation behind a Simulink model based process, the virtualization process and applications of the resulting virtual ECUs. Domain: Critical Transportation Systems Topic: Processes, methods and tools, in particular: virtual engineering and simulation 1. Motivation Since 2010, Renault has established a framework to develop engine control software for Diesel and Gasoline engines [6]. The framework is heavily based on MATLAB/Simulink and the idea of model-based development, which facilitates the carry-over and carry-across of application software between software projects. In the Renault EMS architecture software is composed in to about 20 functions, such as Air System, Combustion etc. A function consists of modules. A module is the smallest testable software unit and contains runnables to be scheduled and executed by the Operating System (Os) of the ECU. The Renault EMS development process includes basically the following steps [5]. 1. Specification of about 200 generic configurable modules per ECU using MATLAB/Simulink. 2. Generation of C code (EMS application software) from all module specifications using MATLAB/Simulink Embedded Coder. 3. MiL (Model in the Loop) test and validation of the resulting executable specifications at module level in a simulated system environment, considering only essential interactions with other modules and system environment. This is essentially a back-to-back test to make sure that the Simulink model of a module and the corresponding production C code show equivalent and intended behaviour. To insure software quality, this step is repeatedly performed with steps 1 and 2, based on the simulation capabilities of MATLAB/Simulink. 4. Configuration of modules to fit to the specific needs of a software project, such as absence or presence of certain components. 5. Integration of generated configured C code and hand-coded platform software (basic software) on supplied target hardware, a real ECU that communicates with other controllers via CAN and other busses. 6. Validation and test of all modules on system level using the real ECU. In contrast to step 3, the interactions of all modules and interactions with the system environment are visible then and subject to testing. For example, the Os runs all scheduled runnables then, not just those of the modules considered to be 'essential' for a module under test. Critical assessment of the above process shows that there is a considerable delay between delivery of a set of specifications to the software project team (at the end of step 3) and system-level tests based on an ECU that runs entire software (step 6). Typical delays are weeks or months.

Explore related subjects

cs.SE

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Dirk Von Wissel, Yohan Jordan, Dirk Von, Wissel Renault, Adrian Dolha, Jakob Mauss. 2018-02-16. Full Virtualization of Renault's Engine Management Software and Application to System Development. https://arxiv.org/abs/1802.06841

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

Is Vibe Coding Safe? Benchmarking Vulnerability of Agent-Generated Code in Real-World Tasks

Vibe coding is a new software development paradigm in which human engineers prompt a large language model (LLM) agent to complete complex coding tasks with little supervision. Although vibe coding is increasingly adopted, is the generated code really safe to deploy in production? To investigate this question, we propose SUSVIBES, a benchmark consisting of 186 feature-request software engineering tasks from real-world open-source projects, for which, human programmers committed vulnerable implementations. We evaluate 12 widely used coding agentic settings with frontier models on the benchmark. Disturbingly, all agents perform poorly in terms of software security. Although 57% of the solutions from SWE-Agent with Claude 4 Sonnet are functionally correct, only 11.8% are secure. Further experiments demonstrate that preliminary security strategies, such as augmenting the feature request with vulnerability hints, cannot mitigate these security issues. Our findings raise serious concerns about the widespread adoption of vibe coding, particularly in security-sensitive applications. The code and dataset are available at https://github.com/LeiLiLab/susvibes. The leaderboard is at https://leililab.github.io/susvibes-leaderboard.

cs.SE↗
arXiv2026

Why3-py: A Tool for Formal Verification of Hypothesis Testing and Meta-Analysis in Python

The reproducibility crisis in scientific research has received widespread recognition, thereby increasing the importance of meta-analyses that integrate statistical analyses from multiple studies. However, statistical methods often have ambiguous and implicit underlying assumptions, which can lead to their erroneous applications and interpretations. To address this issue, we propose a formal verification framework for statistical Python programs. Specifically, we present Why3-py, a Python front-end for the Why3 verification platform that transforms Python code into verification-oriented WhyML representations, addressing the challenges arising from Python's dynamic typing and runtime polymorphism. Furthermore, we extend the StatWhy tool to support the verification of meta-analysis methods. These tools enable meta-analysts to identify overlooked assumptions and misuse of analyses, and to verify the correct use of hypothesis testing and meta-analysis methods in Python code.

cs.SE↗
arXiv2026

What is the Difference Between Me and You? Benchmarking the Quality Gap Between Human-Written and AI-Generated Code

AI coding assistants are becoming co-authors of production software, yet their evaluation centers on functional correctness, leaving open whether their code differs from human code in the quality dimensions dominating lifecycle cost. We compare human-written and AI-generated code at scale: 787,562 function pairs across Python, Java, and C, each human function mined from open-source repositories paired with implementations generated from its docstring by three AI assistants (OpenAI GPT models, DeepSeek-Coder, Qwen2.5-Coder). We characterize structural complexity and statistical naturalness, and map static-analysis findings onto Orthogonal Defect Classification for defects and the Common Weakness Enumeration for vulnerabilities, making authors and languages directly comparable. AI-generated code is structurally compressed and stylistically templated: roughly half the size and branching of human code, clustering apart at the style level. Defect profiles differ in kind: human code concentrates issues of mature codebases, AI code repetitive boilerplate; security is language-dependent, with LLMs producing more, and more severe, findings in Python and Java but fewer high-severity memory-safety findings than humans in C. Once size is controlled for, complexity metrics carry little signal, while naturalness separates authors. Finally, we release CQBench, a benchmark of 27,346 issue-prone tasks with baselines and an evaluation pipeline for quality assurance and security testing.

cs.SE↗