As the automotive industry becomes increasingly crowded, with automakers and suppliers seeking larger stakes in developing future mobility solutions, identifying intelligent innovations for expediting R&D is critical to gaining a competitive edge.
Product development becomes more digitised than ever, introducing opportunities to deploy advanced Artificial Intelligence (AI) software. A recent Forrester survey found that more than two-thirds (67%) of engineering decision-makers felt pressure to adopt AI into their engineering workflows to avoid losing a competitive advantage. And that’s because AI tools offer a vast array of benefits for unlocking faster product development and higher-quality solutions. They expedite time-consuming validation stages and enable simultaneous training of multiple self-learning AI models—which improve and become more valuable when fed with more data and properly put into context by the engineer.
However, identifying the right data—and how to best apply it in the engineering workflow—can be an incredible time-drain. Indeed, data scientists—many of whom aren’t engineering domain experts—don’t possess the understanding of the testing procedures to effectively target existing datasets that can be leveraged by AI, leading to unnecessary time and resource investment.
This means that organisations need to strengthen their data management systems for identifying the useful data, the relevant testing or engineering context, or equip their product development teams with the right tools to efficiently use the data, context and AI software themselves. Only then will AI’s ability to improve product development truly drive efficiency throughout the engineering workflow, and therefore effectively facilitate better product innovation.
To compound the challenge, research has shown that while senior leadership figures understand the potential of AI, a tiny percentage of their engineering teams are using machine learning to perform root cause analysis with historic or even current test data. Indeed, modern physics-based simulation methods or proven hardware-testing procedures from the 2000s are still used throughout the R&D process. Even if an organisation has a data strategy, it can still be difficult for engineers to correctly identify proper historic data from these legacy systems that can be leveraged by the engineers without disrupting existing workflows.
AI holds an enormous amount of potential for helping automotive businesses predict trends and reach meaningful solutions
Ultimately, data and its correct product-/test-related context is crucial to realising the full potential for AI to enhance product development. The accuracy of AI models depends not only on the quantity, but also the quality of the data. Product manufacturers therefore need to apply time and effort to understand the data they possess, consider the complexity of the problem they’re looking to resolve with AI, and the number of datapoints along the process they’ll need to acquire. Advanced tools like AI aren’t intended to replace traditional engineering programmes but to employ valuable existing data to augment the engineering and testing process and ensure accurate, reliable results in less time.
AI holds an enormous amount of potential for helping automotive businesses predict trends and reach meaningful solutions. Consequently, giving engineers the means to understand their data at their fingertips, how best to apply it to their existing workflows, and how to effectively feed it into an AI model, can unlock a faster route to innovation and give the organisation a competitive edge.
No-one knows the value of engineering’s expertise and related data better than engineers themselves, but ultimately, they don’t know what they don’t know. AI can unlock the data’s true value to a business.
The opinions expressed here are those of the author and do not necessarily reflect the positions of Automotive World Ltd.
Oliver J Walter is General Manager of Automotive, Monolith, an artificial intelligence (AI) software provider to the world’s leading engineering teams
The Automotive World Comment column is open to automotive industry decision makers and influencers. If you would like to contribute a Comment article, please contact editorial@automotiveworld.com