CONNECT WITH US

L3 may be late, but automotive revolution rolls on

Nuying Huang, Taipei; Elaine Chen, DIGITIMES Asia 0

Credit: AFP

It's unclear how many still recall the optimism with which 2025 was once hailed—by automakers across Europe, the US, and even China—as the definitive launch year for Level 3 (L3) autonomous vehicles. Yet, as the calendar pages turn, the odds of L3 systems arriving on time—and making a triumphant debut—now seem slim.

However, the broader evolution of automotive electronics and electrification (EE), along with the software-defined vehicle (SDV) revolution, shows no signs of slowing. These innovations are advancing with or without L3 autonomy, hitting the market.

Computex 2025 is poised to serve as a strategic early indicator of this transformation. Once focused squarely on consumer electronics, the annual Taipei event has evolved into a barometer of the auto industry's shift toward integrated EE systems, bringing together global semiconductor giants, IDMs, and ICT leaders. All signs point to a new phase in automotive architecture—one that increasingly embraces domain- and zonal-controller hybrids.

According to supply chain sources, global automakers are transitioning to a hybridized design framework that combines domain (D) and zonal (Z) control systems. The two key forces: the potential for new commercial opportunities and aggressive cost optimization.

Innovation sparks opportunity

Much of the commercial promise driven by SDVs stems from the rise of intelligent connectivity—both inside and outside the vehicle. From smart cockpits to immersive infotainment and advanced user experiences for drivers and passengers alike, the SDV offers a playground for innovation. Not every design will be a hit, and some past experiments have failed to resonate with consumers. But these are viewed as growing pains rather than deterrents.

What's clear is that with a more robust EE architecture, the scope of SDV-driven opportunities expands significantly.

Cost-efficiency gains traction

Industry insiders reveal that most major automakers—including those in Europe, the US, Japan, South Korea, and leading Chinese brands—are leaning toward the D+Z hybrid approach primarily for its cost-to-performance advantage.

This design doesn't just support the SDV roadmap; it enables smarter resource use. By reducing the reliance on copper wiring and components, automakers can lower both vehicle weight and production costs, benefits that resonate strongly across a highly competitive market.

Once production scales, the D+Z hybrid architecture outperforms traditional domain-based systems and legacy ECU frameworks in terms of cost and flexibility. In simpler terms, this hybrid strategy—often referred to as hybrid zonal or full zonal control—offers a modular, building-block approach. It enables efficient resource allocation and scalable upgrades according to vehicle function.

Taiwanese firms in the spotlight

Global automakers are also increasingly turning to Taiwanese suppliers—most notably Foxconn and other electronics giants—for their deep systems integration capabilities. These companies offer a rare blend of cost efficiency and performance, honed over years of collaboration with major auto brands. Their strengths also extend to critical areas such as functional safety and cybersecurity, non-negotiable requirements in the automotive domain.

Autonomy delayed, but not derailed

While 2025 may no longer mark the commercial dawn of L3 autonomy, the momentum behind EE and SDV development is undiminished. The reality is that L3 and beyond depend less on technical readiness and more on the regulatory environments in which they must operate. Progress hinges on the pace at which governments finalize and enforce policies.

Automakers are not waiting. They continue to invest, develop, and refine. Their strategy: be fully prepared, so that once regulations are in place, they can move swiftly to capitalize.

Until then, areas unbound by legal limitations—like EE and SDV innovation—will continue to propel the industry forward.

Article edited by Jack Wu