Search Results

The rapid increase of networked electronic control units in airplanes (Line Replaceable Units or Modules, LRUs/LRMs) and automobiles (ECUs) requires to move from CAN buses to higher performance buses. In aircraft the number of LRUs exceeded 100 in 1990 (B777) and is now ≻5000 (A380). Today, the number of ECUs in some automobiles also exceeds 100. Aircraft industry developed solutions based on standard switched Ethernet (AFDX) and standardized ECUs, called Integrated Modular Avionic units (IMA units) and common remote data concentrators (cRDCs) that are now flying in the Airbus A380 and A400M, the Boeing B787, and are being used in the design of future civil and military aircraft. During the last decade, automotive industry has been pursuing the development of specialized FlexRay bus solutions for automotive control and specialized MOST bus solutions for comfort electronics. However, some automotive companies are now also looking at Ethernet-based solutions.

Networked systems in automobiles and aircraft are designed by groups of people, departments and/or companies of different expertise. Commonly used design methodologies in complex system design partition the work at high level of uncertainty into teams of specialists. These teams have to make assumptions and design decisions without being able to evaluate the impact of their decisions on the overall design. Many design decisions have to be made during integration. This design methodology does not permit optimization at vehicle level and results in very low probabilities of not having critical design errors [1]. Hence it causes high integration costs and redesigns. This paper shows why the current design methodologies cause unverified designs and critical errors, and how system level design automation technologies can be extended to early vehicle level design stages for verifiable vehicle level executable specifications and architectural optimization at vehicle level.

In defining innovative and cost-effective chip sets for future automotive applications, system architects need high-level tools that allow them to rapidly determine the best silicon partitioning for a given application in terms of system performance as well as cost. The tool needs to be flexible, modular, and swift such that the system designer can perform abstract simulation iterations quickly for various functional partitioning scenarios, without requiring excessive computer resources. The tool must also be portable and adaptable to provide a simulation environment suitable to systems- or car-manufacturers for in-depth applications simulation and architecture assessment. The semiconductor component definition process using such a “mission-level” design tool for the automotive application electronic valve will be demonstrated. Methods for the analysis of electronic valve control system architectures using mission-level simulation will be developed.