New Digital Tools Driving Restructuring for Autonomy
Automakers train and transform to better understand the latest digital development tools, virtual design and testing.
The seemingly unending explosions of fast-changing automotive technologies continue to drive massive restructurings for both management staffs and system architectures. Automakers are revamping corporate structures while simultaneously altering vehicle architectures as engineers strive to meet the demands of autonomy, connectivity, electrification and advanced safety. The transformations driven by the escalating impact of electronics were a central focus for automotive executives who addressed the recent dSPACE World Conference, with effects reaching the upper layers of corporate staffs.
At BMW, for instance, new training is endowing every manager with a deeper understanding of electronic technologies and digital design tools. “All managers have to go through a course in electronics and software,” said Alejandro Vukotich, senior vice president, fully automated driving, at BMW AG. “They even need to take a test after each module to ensure that people actually learn what’s in the coursework.”
Even engineers are being retrained. Specialists in mechanics are getting updates in the latest digital technologies, which include simulation and validation, along with fundamentals of digital technologies. Automakers that have added thousands of electronics engineers and programmers in recent decades now are making sure that other specialists understand how digital devices interact with vehicle mechanics.
“One challenge has been with our mechanical-engineering team – they have been training themselves not to be afraid of electronics and software,” said Andy Griffiths, chief engineer, software and system integration and test, Jaguar Land Rover. “Every engineering manager is being trained in key elements of software, test and validation.”
Several speakers noted that the rapidly expanding role of digital design tools is prompting similar restructurings in design, development and test processes. Streamlining processes and ensuring that tests can be repeated and reused are among driving factors for some OEMs and software-generated virtual tests now dominate many strategies.
New methods are crucial
“A year-and-a-half ago, we started a consolidation with vehicle architectures for test; everything needs to fit in this architecture,” said Dr. Peter Oel, head of E/E integration, simulation and test at Volkswagen AG. “If we don’t change our processes, we won’t be successful. This is disruptive in the way we use testing and integration techniques. We won’t buy standard test benches any more. We want to get modular systems and to be able to use best-in-class tools for each job." He added that cloud technology is a critical element of this new test strategy. Using the cloud makes it possible run tests using massive parallelization so that voluminous full-vehicle tests can be run in comparatively short timeframes.
Strategies for enhancing design and validation take many forms. The same technologies that make designs more complex can be used to help engineers better meet users’ needs. For example, JLR is leveraging connectivity and firmware over-the-air (OTA) updates to help engineers speed new features and functions into production, then tweak them after they’ve seen how drivers interact and use the technology.
“Rather than trying to develop fully a complex product where we’re guessing the user requirements, we try to define a minimum product target, then learn from it and add features with FOTA,” said Alex Heslop, director – electrical engineering at Jaguar Land Rover. The volume of data from inside and outside the vehicle continues to soar and the interactions between modules that create and use this information also are multiplying. That puts more emphasis on ensuring that all parties involved in development and testing are working closely towards common goals.
“The big challenge is addressing the holistic complexity of the car’s functions, there are a lot of concerns inward and outward, and they’re highly dynamic,” said Dr. Chen Ma, Product Owner, Volkswagen AG. “You need everyone to work towards one goal. That’s not easy, because everyone normally works towards their own goals. This is now a highly interdisciplinary domain – companies need to find a level of working together that is a golden point.”
Buying into AI
The deluge of data used by vehicles and the need to accurately predict how myriad automotive systems interact makes predictive design analysis a critical tool for vehicle development. At its first worldwide user conference, held in Munich, Germany, dSPACE said the company has transformed to address new demands. Martin Goetzeler, dSpace’s CEO, detailed the company’s expansion to a full-service provider during the past two years. “We’ve become an end-to-end simulation and validation provider,” Goetzeler said. “We’ve extended our hardware-in-the-loop capabilities to include software simulation, software-in-the-loop and cloud simulation as a service.”
Over the past two years, manpower has increased by more than 30% to surpass 1,800 and new corporate R&D and test facilities opened. In June 2019, dSpace acquired three-year-old startup Understand.ai to address the burgeoning role of artificial intelligence. This broad restructuring sets the stage for years of double-digit growth, Goetzeler predicted. While the changes in the past two years have been significant, the restructuring at dSPACE and many other automotive companies is far from finished. “We’re in the middle of a transformation driven by e-mobility and autonomy,” Goetzeler said. “We need to continue altering our company and work with new partners.”