Mixed-Criticality Systems

Mixed-Criticality Systems (MCS) refer to embedded systems where tasks of varying criticality levels—ranging from safety-critical to non-safety-critical—are executed on the same hardware platform. This design challenge requires ensuring that high-criticality tasks always meet their timing and reliability requirements, even under heavy system load. The concept originated in aerospace but has become central to automotive cybersecurity and functional safety. According to ISO 26226 and ISO 26262 standards, the system must be designed to prevent interference between tasks of different ASIL levels. This is achieved through mechanisms like memory protection, temporal isolation, and priority-based scheduling. In a risk management context, MCS represents a critical control measure to prevent systemic failures caused by resource contention. The primary objective is to be able to be certain of the execution of safety-critical functions, even when the system is operating at its limit. This allows for higher-performance computing in vehicles without compromising passenger safety, which is essential for the advancement of ADAS and autonomous driving technologies.

Implementation of MCS in enterprise risk management typically follows three stages: Classification, Isolation, and Verification. First, the enterprise must categorize all software functions by their Automotive Safety Integrity Level (ASIL) according to ISO 26262. Second, the system architecture must be designed with hardware-enforced isolation, such as using a hypervisor or MPU to prevent low-criticality tasks from accessing or corrupting high-criticality memory regions. Third, dynamic scheduling algorithms must be implemented to ensure high-priority tasks always meet their deadlines. For example, a European-based automotive supplier implemented MCS in their next-generation ECU, which integrated both infotainment and ADAS functions. This consolidation reduced the number of ECUs by 30%, lowering both-end costs and weight. The company reported a 20% reduction in safety-related incidents during the first year of production due to better-defined task priorities. This approach aligns with the EU's TISAX cybersecurity standard, which requires clear evidence of functional segregation and risk-adjusted resource allocation.

Taiwanese enterprises face three primary challenges: technical talent shortage, high verification costs, and regulatory uncertainty. The first challenge—technical talent—can be addressed by investing in upskilling existing engineers in RTOS principles and safety standards like ISO 26262. The second challenge involves the complexity of verifying MCS; traditional testing methods are insufficient to guarantee safety in all load scenarios. Companies should adopt model-based testing and formal verification tools to provide mathematical certainty of task isolation. The third challenge is the evolving regulatory landscape, including the EU AI Act and Taiwan's emerging cybersecurity regulations for connected vehicles. To overcome this, enterprises must be closely closely monitoring international standards and adopt a modular design that allows for updates without re-certifying the entire system. A phased approach—starting with small-scale integration before full-system deployment—is recommended to manage both cost and regulatory risk effectively. The estimated time-to-value for a well-managed MCS implementation is typically 12 to 18 months, with significant ROI realized through reduced hardware costs and lower warranty-related risks.

Winners Consulting Services Co., Ltd. specializes in Mixed-Criticality Systems for Taiwan enterprises, delivering compliant management systems within 90 days. Free consultation: https://winners.com.tw/contact