In a featured session at Automotive IQ's online Autonomous Vehicle event, David Doria, Director of Engineering – Automated Driving at Magna International, moderated an engaging panel discussion with Ahmed Abutabl, Engineering Director and Head of Global Software Engineering – Body Control at FORVIA HELLA, and Sergey Malygin, CEO of SODA.Auto.
Together, the panel explored the technologies, architectures, and engineering practices shaping the future of autonomous vehicles and software-defined vehicles. Discussions ranged from AI scalability and zonal architectures to validation strategies, simulation, and the future role of V2X connectivity.
Below are the top 10 takeaways from the discussion.
To watch the full session, click here.
1. Scaling AI Is Now a Bigger Challenge Than Building It
The industry has largely demonstrated the potential of AI for autonomous driving. The next challenge is scaling these capabilities into production vehicles through robust engineering processes, validation frameworks, and deployment strategies.
2. Zonal Architectures Are Reshaping Vehicle Design
Automakers are increasingly adopting zonal architectures to reduce wiring complexity, vehicle weight, and hardware redundancy. This transition enables more efficient centralized compute platforms while providing a stronger foundation for software-defined vehicle development.
3. SDVs Depend on Centralized Compute
Software-defined vehicles require centralized computing architectures capable of processing vast amounts of sensor data while supporting over-the-air updates, feature upgrades, and subscription-based services throughout the vehicle lifecycle.
4. Autonomous Driving Systems Are Moving Toward End-to-End AI
The evolution from traditional sensor fusion approaches to end-to-end neural networks reflects a broader industry shift toward AI systems that can better interpret complex driving environments and address challenging edge cases.
5. The Industry Is Preparing for Higher Levels of Autonomy
While Level 2 systems dominate today's market, significant development efforts are focused on overcoming operational design domain (ODD) limitations and enabling more capable autonomous driving systems.
6. The Boundary Between L2 and L4 Could Blur
Rapid advancements in AI performance and computing power may accelerate the convergence of technologies currently used in assisted driving and fully autonomous systems, potentially shortening the path to higher levels of automation.
7. Testing Strategies Must Evolve Alongside System Complexity
As vehicles become increasingly software-centric and AI-driven, companies must move beyond siloed testing approaches and adopt comprehensive validation strategies that integrate software, hardware, simulation, and vehicle-level testing.
8. Simulation Will Play an Increasingly Important Role
Simulation remains one of the most powerful tools for rapid prototyping and validation, enabling developers to test scenarios at scale. However, limitations remain in accurately modelling certain sensor technologies, reinforcing the need for a balanced approach that combines virtual and real-world testing.
9. V2X Could Unlock the Next Wave of Vehicle Intelligence
V2X communication has the potential to enhance situational awareness, improve decision-making, and enable new services by allowing vehicles to leverage infrastructure data, connected ecosystems, and potentially off-board computing resources.
10. Trust, Transparency, and Traceability Are Becoming Strategic Priorities
As autonomous systems grow more sophisticated, success will depend not only on technical performance but also on clear safety metrics, transparent validation processes, and development workflows that provide full traceability across increasingly interconnected vehicle systems.
These topics, and more, will be explored at Automotive IQ's SDV & AV Technology 2026 conferences.
Find out more about SDV & AV Technology Silicon Valley 2026 here.
Find out more about SDV & AV Technology Europe 2026 here.