Bachelor's and Master's from Xi'an Jiaotong University; Senior Engineer; 15 years of automotive electronic system development experience; First-author SCI publications, 20+ core patents; focused on complex electronic control system architecture methodologies, boosting enterprise digital R&D efficiency, and empowering industry engineers.
This article explains the Arcadia four-layer architecture method (Operational, System, Logical, Physical), detailing each layer's purpose, the risks of jumping straight to physical design, and how bidirectional traceability across layers controls rework costs and aligns team communication.
This article presents a six-step methodology for creating system concept documents based on the Arcadia framework, distinguishing concepts from requirements and architecture through operational analysis, boundary definition, black-box capabilities, trade studies, and baseline freezing with concrete automotive examples.
This article distinguishes architecture from architecture diagrams and outlines eight interconnected capabilities every architect needs: abstraction to identify key attributes, focus to prioritize constraints, decomposition by design decisions, boundaries to assign ownership, contracts to decouple implementations, trade-offs as explicit prioritization, adapting to constraints as design inputs, and evolvability as the compound result of the previous seven.
This article explains how a single JSON message can carry both business data and control signals, and why distinguishing them by purpose — not format — is critical for correct Data Flow Diagrams and choosing the right UML diagrams for control flow.
This article reveals how ambiguous concept boundaries—not technical mistakes—drive architecture costs, using automotive examples to show how clarifying intension, extension, and system concepts enables clear responsibilities, shared language, risk analysis, and organizational knowledge.
Cognitive science shows the human brain can only deeply focus on two tasks at once due to limited working memory, attention residue, and long context-switch recovery; by externalizing judgment rules into AI agents, knowledge workers can offload mature tasks to run dozens in parallel while maintaining two front-line efforts.
Contrasting DeepMind's GNoME discovering 2.2 million crystal structures with Apple's finding that LLMs fail simple math when irrelevant details are added, the article argues AI excels at generating candidates but lacks true reasoning, so humans must define value, set boundaries, and take responsibility while using AI for rule drafting, gap checking, translation, and simulation.
The article analyzes two contradictory studies on AI coding assistants: one showing a 55% speedup for well-defined tasks, another showing a 19% slowdown in complex real-world codebases. It argues the difference lies in whether tasks have explicit, codified rules versus tacit knowledge, and advises extracting implicit rules into executable constraints to harness AI effectively.
Industry experts discuss how automotive electronic architecture uses layered function-requirement decomposition, logical architecture, and bidirectional traceability to verify designs, comparing tools like Capella and Preevision, and analyzing a high-voltage power-down fault caused by missing indirect function-chain analysis.
This article explores the concept of functions from their mathematical definition through programming language implementation, detailing key elements, stack-based execution mechanics in C, and three perspectives: functions as reusable abstraction units, control flow management tools via call stacks and callbacks, and operators in dataflow architecture pipelines.
