Despite joint ventures and partnerships, sharing always has been a difficult action across the auto industry, given the organizational silos and inter-partner rivalries often standing in the way. But there is one shared area in which everyone can benefit: the ever-widening application of digital integration design tools (DIDT).
Experts tell SAE International there is increased urgency to speed the development and application of DIDT within and between companies. They cite the time and cost of developing highly integrated and complex electric-vehicle propulsion technologies, and automated-driving systems, as examples. Seeking common ground, a group of two dozen senior design engineers representing vehicle OEMs, Tier 1s, consultancies and academics, recently met privately in the U.K. to sort out the challenges and hurdles facing digital design tool integration. Their workshop generated many useful insights and potential solutions, shared with the author and highlighted below. “The automotive industry is a long way from having the digital tools we need to meet the intense cost and timescale pressures ahead,” said Prof. Sam Akehurst at the newly-established Institute for Advanced Automotive Propulsion Systems (IAAPS), part of the University of Bath. “Even more concerning is that if those tools were available, many companies would find that current organizational silos would prevent their adoption.”
When IAAPS formally opens in 2021, it will be the first institution in the U.K. to be created specifically to develop the tools and skills needed to accelerate the development of digital engineering, research and development — and to facilitate more efficient innovation at companies of all sizes.
“It’s not just pressure from the tremendous fragmentation of powertrain technologies or the acceleration of new areas like ADAS, or even new ways of working in response to Covid-19,” explained Akehurst, who, as professor of Advanced Powertrain Systems at Bath, is a leading campaigner for greater application of digital design tools. “The focus of digital tools has traditionally been on the analysis of individual systems,” he said. “The biggest challenge — and one of the most interesting opportunities — is to step this up so the technique can be used to inform decisions that cover several related systems and the many options for tight physical and functional integration.”
The group discussion concluded that the industry is a long way from having the digital tools needed to meet intense cost and timescale pressures, said Akehurst.
The University of Bath will use the results of the workshop to build a digital development roadmap, together with industry partners and the U.K. government’s Automotive Propulsion Center (APC), to be available to organizations worldwide to help guide the development of a new generation of design-supporting digital tools.
Akehurst noted there already are verification tools to reduce physical testing, some key areas of which are not even in development. “There is also the more recent area of high-level design support that can make a substantial difference to vehicle cost and efficiency, particularly from improved decision making at the concept stage,” he said.
Alex Read, a senior director at Siemens Digital Industries Software, believes that each of these two activities are philosophically different. “While it is right that correlation with the real world is a priority, we need to realize that in some important areas where digital tools can add value, it’s an orientation that is needed, not absolute fidelity,” he said, citing the time required to assess design options as an example. “A key benefit of digital prototyping is that the cost of evaluating alternatives will be low,” Read said. “The next generation of digital tools must help us test more options, across a wider spectrum of possible solutions, to move from single numbers to hundreds or even thousands.”
It’s an area where digital system suppliers are already making significant progress, with several companies offering end-to-end toolsets that join up the capabilities of their more focused offerings. But many vehicle manufacturers and consultancies believe these don’t answer enough major questions. As a result, they have developed their own proprietary decision-support tools.
While this may answer an immediate need for these companies, adoption is patchy, according to Akehurst. And too little interoperability means opportunities for collaboration among industry partners cannot be fully exploited. The IAAPS sees this as increasingly important, especially in the post-COVID, resource-constrained future. Akehurst adding that it can’t be just about accelerating innovation: “It also has to be about specialists from a vast range of companies and academia,” he asserted, “working much more closely together to achieve very high levels of optimization.”
Siemens’ Read agreed. “The workshop confirmed the need to change cultures, teach new collaboration skills, and facilitate system-level collaborations, with digital tools a vital part of that journey.”
Another important area to which increased collaboration looks set to contribute is the integration of incumbent standalone (or only physically integrated) systems into single, highly optimized “super-systems,” as IAAPS terms it. For example, an EV’s transmission, motor and power electronics are typically developed separately, often by different suppliers. Designing them as a single, highly optimized super-system can reduce cost, improve packaging and greatly increase system and vehicle efficiency.
But acknowledging this and achieving required results are two very different things. A new generation of powerful digital design tools, with an embedded facility for secure, remote collaboration, could make optimization “faster, cheaper and much more intense,” Akehurst said.
So far, though, it is not happening at the rate required. One of the main reasons for this, explained Prof. Chris Brace, IAAPS Academic Director, is the historic functional silos found in many large companies. “We all agree that a new generation of decision support and super-system optimization tools is needed,” Brace said, “but extracting the benefits from them needs a more open internal structure, too. Functional silos represent one of the biggest blocks to accelerating development of this much-needed new approach.”
Goodbye physical testing
Another opportunity is offered by the integration of Big Data, possibly collected from test vehicles but also from the in-service fleet via the cloud. IAAPS has evolved new versions of standard statistical tools to help extract trends from large, complex datasets and believes this could be used alongside prognostics to continually improve decision making. “There is so much data available, we can make simulations much closer to the real world and real driving,” stated Akehurst. “Equally importantly, we can use this data to help us with design decisions such as how to accommodate edge cases in component life targets.”
Ricardo’s chief engineer for Technology Strategy, Joshua Dalby, supported this view. He said that a related area where simulation could be contributing more value is via the investigation of edge cases, as these are difficult to recreate reliably and efficiently in a physical test environment. “These areas offer great potential for saving cost and weight, but to avoid costly in-service failures, designers need a more detailed understanding than can be delivered economically by physical testing,” Dalby offered. “Simulation can explore more options, more cheaply and potentially in days rather than weeks.”
Any design change or new approach — especially in an edge case — has potential to increase risk for vehicle OEMs. New simulation tools provide the ability to assess risks introduced by these changes very quickly, noted Akehurst, allowing more options to be evaluated and better decisions to be made. “I see a lot of potential for further development in these areas,” he said.
What of the contentious issue regarding simulation completely replacing physical testing? IAAPS Engineering Director, Prof. Rob Oliver, believes it is unlikely — at least for the foreseeable future. “The workshop group agreed that systems are evolving too quickly to keep simulation sufficiently far ahead of technology developments,” Oliver said. “New battery chemistries represent a great example. In other areas, like combustion, we are years from having enough computing power and enough fidelity in the algorithms to remove the need for physical experiments.”
He added that, paradoxically, accelerating the application of digital engineering needs a realistic understanding of the technique’s limitations and a strong focus on physical testing to explore and resolve areas of poor correlation. This is critically dependent on all groups working to common standards so that their data is accessible, verifiable and understandable to others in an organization.
“The strategy must be implemented at a senior level, where organizational culture can be changed,” Oliver stressed. “The workshop opened with many delegates expecting to conclude that simulation technology is the roadblock, but we finished by concluding it is the approach of senior management that needs to change.”Continue reading »