Agent Factors Engineering
Xuetao Wei · 2025
With the rapid evolution of artificial intelligence technology, AI agents have evolved from basic task executors to sophisticated agents with complex decision-making capabilities, expanding their applications from schedule management and routine document processing into mission-critical domains such as autonomous driving, precision medicine, and intelligent investment advisory. With the significant enhancement of intelligence levels and autonomous decision-making capabilities in agents, while bringing about an efficiency revolution, the systemic interoperability risks, task execution efficiency bottlenecks, and absence of safety mechanisms exposed during their deep integration into core domains of human society have increasingly drawn attention from both academic and industry communities. In this position paper, we introduce the concept of Agent Factors Engineering(AFE) 1 and discuss research challenges arising from the interplay between AI agents and engineering systems, such as interoperability, efficiency and safety. AFE aims to promote the synergy between AI agents and engineering systems, seeking theoretical foundations and technological pathways for improving their productivity and social good.