Program Initiation/Switching TRIZ Cases
- Multi-Core Vehicle Control for Interrupt Response and Task Isolation
Separate cores handle time-sliced control tasks and interrupts to avoid software interference while improving vehicle control response.
- Complexity-Scored Agent Prediction Scheduling for AV Planning
Complexity and priority scoring allocate limited AV compute to nearby agents, preserving timely prediction data for trajectory planning.
- Networked ECU Task Scheduling for Deadline Verification
Dynamic ECU task scheduling switches execution patterns by processing load to meet deadlines while limiting verification complexity.
- Vehicle Domain Controller Platform for Real-Time SOA Data Handling
A layered SOA vehicle platform keeps critical data inside domain controllers to improve real-time control, security, and maintenance.
- Vehicle Control Unit Scheduling With Cross-Core Backup Threads
Cross-core backup threads and flag monitoring keep vehicle control tasks on time under AUTOSAR adaptive scheduling limits.
- Adaptive ML Task Batching for Latency-Energy Tradeoffs
Dynamic batch sizing and execution timing use queue-state cost modeling to balance ML inference latency against energy consumption.
- Sensor Data Processing Architecture for Low-Storage Multi-Task Recognition
A split processor and selector architecture shares feature extraction across tasks to cut software storage while preserving recognition performance.
- Autonomous Vehicle Fallback Task Triggers for Safe Self-Recovery
Predefined fallback tasks and trigger conditions let autonomous vehicles reroute to maintenance, charging, or refueling before failures cause stranding.
- Containerized Control Architecture for Scalable Processor Switching
A brokered container architecture lets control applications move across processors without recompiling, improving scalability and real-time validation.
- Distributed Datacenter Power Control for Renewable Curtailment
A flexible datacenter modulates behind-the-meter load to absorb curtailed wind and solar power while supporting grid stability.
- Vehicle Control Partition Switching for Dynamic Function Updates
Partitioned function groups let a vehicle controller switch condition-specific software during operation without interfering with active control.
- Constraint-Aware Task Planning for Dynamic Obstacle Navigation
Stochastic optimization turns sensed constraints into costs, enabling real-time task planning around static and moving obstacles without retraining.
- Logic-on-Logic Die Stacking for AI Bandwidth and Thermal Control
By placing the compute die above memory, this case boosts inter-die bandwidth while improving heat removal and reducing TSV burden.
- Vehicle Task Trigger Control for Processor Failure Timing
Predetermined task timing and malfunction detection trigger fallback actions when vehicle processors miss deadlines, preserving stability and safety.
- Multi-SoC Fail-Safe Orchestration for Autonomous Driving Workloads
Critical workloads are migrated across a local SoC cluster so autonomous driving functions stay available after node failure while lowering silicon cost.
- Distributed Vehicle Data Processing for ECU Throughput Bottlenecks
Idle-time-based task offloading across ECUs and remote processors boosts vehicle software throughput while reducing reset and data corruption risk.
- NoC Arbitration Priority Control for Fair Shared Resource Access
Dynamic NoC arbitration updates request priority from pending queue levels to curb greedy initiators, reduce timing issues, and improve fairness.
- Sensor Data Timing Around Windshield Wiper Interference
By delaying sensor acquisition outside wiper activity, this case preserves camera data quality for more reliable vehicle travel control.
- Decentralized ECU Task Offloading for Vehicle Software Throughput
Idle ECU capacity is used to offload non-safety-critical tasks, improving vehicle software throughput without harming performance or safety.
- Programmable Atomic Unit Thread Self-Scheduling for Memory Latency
Thread IDs let a programmable atomic unit reschedule stalled threads around memory requests, cutting hazard-tracking complexity and die area.
- Data Processing Network Execution for Low-Latency Reproducibility
Combining data-driven and time-driven module execution cuts latency while keeping multi-core processing reproducible for automated driving.