소프트웨어 정의 자동차(SDV)가 가속화함에 따라 진화하고 있는 존(zone) 아키텍처에 대해 마우저일렉트로닉스(Mouser Electronics)의 디비야가리카파티(Divya Garikapati)가 이야기한다.
“Zone Architecture Leads SDV Progress”
Easy SW update and expansion, centralized data fusion and unified optimization
ZCU's fast data analysis is important for real-time decision-making and rapid response
As software-defined vehicles (SDVs) accelerate, zone architectures are also evolving. Modular architectures that are dedicated to specific functions have the advantage of increasing efficiency and facilitating expansion with SDVs. Although there are fundamental challenges associated with implementing zone architectures, it is expected that this architecture will lead SDV advancements in the future. Zone architectures will enable safer, more efficient, and more personalized transportation.
■ Benefits of Zone Architecture with SDV
Traditional automotive architectures rely on centralized electronic control units and complex wiring, which fundamentally limits scalability and performance. Zone architecture, on the other hand, breaks down monolithic structures and divides automotive electronic systems into specialized modular zones. Each zone contains a dedicated high-performance computing unit (HPCU) that runs software tailored to its function and maximizes performance. Communication between zones uses a standardized protocol, eliminating the need for cumbersome wiring and enabling smooth data exchange.
Zone architecture is scalable, as existing zones can be updated with software, thus reducing development time and complexity. Seamless data flow between zones allows for central data fusion and unified optimization. This results in a smoother driving experience, improved energy management, and better overall vehicle performance. This software-oriented approach also allows for remote updates and feature additions, thus significantly reducing development schedules and enabling seamless integration of new technologies. Central control and high-speed communications enable rapid and coordinated responses to critical situations. Furthermore, the inherent redundancy and fault tolerance of zone architecture ensures continued operation even if one zone fails. This increases safety and provides a reliable driving experience.
■ Challenges for implementing zone architecture
Zone architecture is expected to revolutionize SDV, but it also presents significant challenges for implementation. Integrating multiple zone control units (ZCUs), sensors, and software across multiple zones requires a high level of expertise and intensive testing to ensure seamless communication. Cost and development schedule are other challenges. Zone architecture requires significant up-front investment and a long development schedule compared to traditional architectures due to its complexity. Continuous software updates across multiple zones further increase resource requirements.
But the most important challenge is safety and security. The distributed nature of the zone architecture increases the attack surface, requiring robust cybersecurity measures, fault-tolerant mechanisms, and clear human-machine interaction (HMI) protocols. Evolving regulatory frameworks and public trust in relation to safety, security, and ethics issues create yet another layer of complexity that is critical to unlocking the full potential of zone architecture with SDV. Fortunately, a fully centralized architecture offers a path forward to address these challenges.
■ Technologies that enable zone architecture
Each ZCU includes a high level of processing performance and HPCU, enabling the unit to handle complex calculations required for autonomous control. For real-time decision-making and rapid response, the ability of the ZCU to analyze data quickly and efficiently is important. Equally important for zone operation is the cognitive ability using advanced cameras, lidar, and radar systems. This ability allows the unit to understand the situation, respond to changing situations in each zone, and make intelligent decisions. This enables a safer driving experience.
Zone architecture incorporates fundamental redundancy in terms of failure. Backup devices and systems act as a safety net, allowing continued operation even if one part fails. This increases passenger safety and minimizes the risk of accidents. Real-time operating systems are another key technology in zone architecture, enabling lightning-fast task execution with ZCUs. Therefore, each zone can respond immediately to sensor data and actuator control, and take quick and intelligent actions based on constantly changing conditions.
Artificial Intelligence (AI) and Machine Learning (ML) are two other important technologies that are emerging as zone architectures. AI and ML algorithms are important for sensor fusion, decision making, and zone-level control. By learning and adapting in real time, each zone enhances its autonomous driving capabilities. Therefore, it improves its ability to cope and adapt on the road. High-speed Ethernet and innovative wireless technologies create a high-bandwidth data highway between the zones and the central system.
Because SDVs do not operate in isolation, communication technologies such as vehicle-to-everything (V2X), vehicle-to-vehicle (V2V), and vehicle-to-infrastructure (V2I) enable SDVs to communicate with the world around them. This collaborative network enhances situational awareness and enables collaborative decision-making between zones and with other vehicles and infrastructure.
Of course, there is still a lot of room for improvement in all these aspects. Exactly what is needed will vary from company to company, vehicle to vehicle, and architecture to architecture. But there is much work to be done in terms of sensor enhancements and ML algorithms. For example, rapidly advancing quantum computing will enable optimization and simulation testing of complex algorithms. Many other challenges remain, including interoperability between devices from different manufacturers, optimizing communication in complex systems, and addressing cybersecurity vulnerabilities. Fortunately, solutions are also on the way. Rapid technological advances offer promising prospects, especially when it comes to centralized architectures.
■ Future Opportunities for Zone Architecture with SDV
As AI, communication technology, and software flexibility advance in the future, the zone architecture will become more intelligent and responsive. With highly scalable hardware and software as a solid foundation, powerful ZCUs will execute complex algorithms and optimize functions in real time. Next-generation technologies like 6G will enable seamless data exchange, providing cars with the information they need quickly and allowing zones to make quick and efficient decisions. Zone AI agents will continually learn and adapt, improving their ability to perceive and control situations. Sophisticated sensor fusion technologies will provide better situational awareness, and transparency and cybersecurity measures in explainable AI (XAI) decision-making will build trust and protect systems.
These advances will enable exciting things in the future: improved cognitive and communication capabilities to reduce accidents and streamline traffic flow, enable personalized driving experiences, and enable human-oriented journeys through external displays and V2X communications. Regulatory frameworks, cybersecurity, and public acceptance remain important challenges. Transparent education and ethical considerations will be needed to overcome these challenges.
■ Conclusion
Zone architecture will play a key role in the success of future SDVs. It will enable vehicles to respond to changing needs with exceptional flexibility and scalability, paving the way for safer, more efficient, and more personalized transportation.
※ About the author
Divya Garikapati is the Director of Autonomous Driving Systems, contributing to the realization of future mobility by leveraging her industry-accumulated knowledge. She has been heavily involved in IEEE and SAE standards and has published numerous articles. He is in charge of level 4 and 5 autonomous driving systems at Woven by Toyota. He has extensive experience in system architecture, functional safety, and model-based system engineering. He holds a master's degree from the University of Michigan and is actively involved in mentoring with the organization 'Women in Autonomy'.