Pragmatic strategies for engineering artificial intelligence into real-world products

Product engineering teams are ramping up artificial intelligence investments cautiously, prioritizing verification, human oversight, and measurable real-world outcomes such as sustainability and quality over rapid transformation.

Artificial intelligence is moving from purely digital applications into physical products such as cars, home appliances, and life-sustaining medical devices, where errors can have serious real-world consequences. Product engineers are turning to artificial intelligence to enhance, validate, and streamline design, but they are doing so along a disciplined and pragmatic trajectory that reflects their responsibility for safety-critical systems. The central tension is how to capture the value of artificial intelligence without undermining product integrity in domains where failures can affect lives, trigger recalls, or cause structural damage.

New research based on a survey of 300 respondents and in-depth interviews with senior technology executives and experts shows that verification, governance, and explicit human accountability are treated as mandatory conditions for deploying artificial intelligence in product engineering. When artificial intelligence systems directly inform physical designs, embedded systems, or manufacturing decisions that become fixed at release, any resulting product failures create risks that cannot be rolled back. In response, engineering organizations are adopting layered artificial intelligence architectures with distinct trust thresholds, rather than relying on broad, general-purpose deployments, to ensure that higher risk decisions receive stricter oversight.

Investment plans reflect this cautious posture. Nine in ten product engineering leaders plan to increase investment in artificial intelligence in the next one to two years, but the growth is modest and structured. The highest proportion of respondents (45%) plan to increase investment by up to 25%, while nearly a third favor a 26% to 50% boost, and just 15% plan a bigger step change between 51% and 100%. Predictive analytics and artificial-intelligence-powered simulation and validation are the top near-term priorities because they offer clear feedback loops, support regulatory approval, and help prove return on investment, enabling organizations to build gradual trust in the tools. Measurable outcomes such as sustainability and product quality rank above time-to-market, innovation, and internal metrics like cost reduction or workforce satisfaction, with real-world signals including defect rates and emissions profiles carrying more weight than internal engineering dashboards.

57

Impact Score

Building a strong data infrastructure for artificial intelligence agents

Enterprises are rapidly experimenting with agentic artificial intelligence, but most struggle to scale because their data architectures lack the business context and trust needed for reliable outcomes. A semantic, business-aware data layer is emerging as the critical foundation for effective agents that work alongside, not instead of, existing software systems.

Imec launches European university consortium for next generation chips

Imec is creating a new consortium with 26 European university groups to push semiconductor research beyond traditional CMOS scaling, focusing on chip design automation and advanced architectures. The collaboration will use Imec’s NanoIC pilot line to accelerate transfer of academic concepts into industry-ready technologies.

Samsung to deploy 2 nm process for HBM4E base die

Samsung plans to manufacture the base die of its next generation HBM4E memory on a 2 nm process, aiming to boost performance and efficiency while tightening its grip on the high bandwidth memory market.

Contact Us

Got questions? Use the form to contact us.

Contact Form

Clicking next sends a verification code to your email. After verifying, you can enter your message.