ADAS Virtual Prototyping with the OpenMETA Toolchain

TitleADAS Virtual Prototyping with the OpenMETA Toolchain
Publication TypeConference Paper
Year of Publication2016
AuthorsEisele, S., M. Yamaura, N. Arechiga, S. Shiraishi, J. Hite, J. Scott, S. Neema, and T. Bapty
Conference NameSAE 2016 World Congress & Exhibition
Date Published04/2016
PublisherSAE International
Conference LocationDetroit, Michigan, USA

Complex systems, such as modern advanced driver assistance systems (ADAS), consist of many interacting components. The number of options promises considerable flexibility for configuring systems with many cost-performance-value tradeoffs; however the potential unique configurations are exponentially many prohibiting a build-test-fix approach. Instead, engineering analysis tools for rapid design-space navigation and analysis can be applied to find feasible options and evaluate their potential for correct system behavior and performance subject to functional requirements.

The OpenMETA toolchain is a component-based, design space creation and analysis tool for rapidly defining and analyzing systems with large variability and cross-domain requirements. The tool supports the creation of compositional, multi-domain components, based on a user-defined ontology, which captures the behavior and structure of components and the allowable interfaces. Design spaces in OpenMETA allow product families to be defined in a single model, with component/subsystem alternatives and parametric variation. Using this system design space, OpenMETA then enables analysis of the system, via composition of the system and environment/scenario models into engineering tools, and executing simulations to compute metrics.

System models can be created and executed in many abstractions based on the required accuracy, phenomena, and execution speed. This paper explores use cases from simulations with high fidelity components, to a gamified environment using Unity with a simple model of vehicle physics. This allows for user-in-the-loop analysis of controllers and components. This approach benefits ADAS by allowing for rapid prototyping across an array of candidate designs while evaluating the requirements of the vehicle at the appropriate fidelity level.