Page 16

EETE SEP 2015

automotive Automotive electronics complexity at tipping point, study warns By Christoph Hammerschmidt Features and functions are added at rapid pace to new cars. Typically, the E/E engineers add a dedicated electronic control unit (ECU) for every new function. This approach is hitting its limits, because it increases the complexity. Consultancy Roland Berger sounds the alarm. Adding a new ECU for new features is no longer sustainable, the study says. Dedicated processors, memories and other electronic components for new features increases cost and architecture complexity, says Thomas Wendt, Senior Partner in Roland Berger’s North American Automotive Practice. The solution he suggests is module consolidation. This approach would leverage modern technologies to add speed and flexibility to vehicle electronic architectures, while saving cost. The consultancy estimates at average $175 per vehicle for cockpit electronics. “All major automotive trends today, from improved cockpit electronics to new ADAS features, are largely enabled by advanced electronics systems,” says Wendt. “OEMs will not be able to keep up with consumer’s expectations, both in terms of quality and price, if they continue to add ECUs every time they want to add a new feature. A ‘blank sheet’ approach to electronic architecture design is needed.” Module consolidation is a technical solution leveraging modern, multicore processing technologies to operate multiple ECUs which all traditionally had their own processors. In a multicore solution, these ECUs retain dedicated processing space, usually in the form of their own core in the processor. However, a number of redundant components are eliminated, including housings, power supplies, wire mounts and harnesses, as well as the processors themselves, all saving cost. Additionally, ECUs communicate within the processor itself instead of communicating over a network such as the CAN bus; increasing speed and reducing complexity. The study quantifies the cost advantages of module consolidation from the perspective of an OEM. Taking a sample set of cockpit electronics, Roland Berger conducted a total cost of ownership (TCO) analysis, comparing the cost of independent ECUs to the cost of a consolidation solution running on a multicore processor with the same feature and function set. The result was a TCO advantage of $175 per vehicle, including direct piece price savings which are just the “tip of the iceberg.” The savings identified also include indirect, yet quantifiable, advantages of consolidation such as weight savings. Despite the clear advantages of module consolidation, OEMs have been slow to adopt the solution. This is largely due to safety and security concerns related to running multiple control functions on the same processor. While there is some merit to this concern, advanced suppliers have already developed a solution to this issue, known as “hardware virtualization.” Due to a clear need and availability of a solution it is now time for the automotive community to re-think legacy E/E architectures and adopt consolidated module solutions. The first movers have an opportunity to capture a tremendous amount of value, both through savings and the ability to offer a superior product to end consumers. These findings are in line with a discussion started some time ago in the European automotive industry. In particular carmaker BMW is advocating a strategy of consolidation and has presented its concept in numerous engineering meetings and congresses. BMW favours a concept of “Domain Controllers” which assume multiple tasks associated to the respective domain, i.e. chassis, body, powertrain or infotainment. In the meantime, the idea of virtualization has also reached the automotive industry. The Genivi infotainment operating system will soon enable virtual machines in cars; and automotive supplier Continental has officially introduced a product concept based on virtualization. ‘Test city’ opens for autonomous cars By Rich Pell The University of Michigan (Ann Arbor, MI) has officially opened a simulated city that will be used for developing and testing driverless and connected vehicles. Located on the university’s campus, the $10 million, 32-acre outdoor environment - called Mcity - features life-sized building facades, various surfaced roads, intersections, signs, and even simulated pedestrians. The environment is designed to allow automakers and others in the industry a controlled environment in which to test connected and automated vehicle technologies. In an attempt to reflect real-world conditions as much as possible, Mcity even includes details such as faded lane markings and partially obscured signs. In addition to testing the physical navigational capabilities of connected vehicles, the environment includes capabilities for testing car-tocar and car-to-infrastructure communications. First announced in May of last year, Mcity was designed and developed by the University of Michigan’s Mobility Transformation Center in partnership with the Michigan Department of Transportation. General Motors, Ford, Bosch, Honda, Nissan and Toyota are among companies that also helped back the project. “There are many challenges ahead as automated vehicles are increasingly deployed on real roadways,” says Peter Sweatman, director of the U-M Mobility Transformation Center. “Mcity is a safe, controlled, and realistic environment where we are going to figure out how the incredible potential of connected and automated vehicles can be realized quickly, efficiently and safely.” 14 Electronic Engineering Times Europe September 2015 www.electronics-eetimes.com


EETE SEP 2015
To see the actual publication please follow the link above