Although the modelling of interactions has long been at the core of socio-technical systems theory, and is a key for understanding resilience, there is a lack of a holistic taxonomy of interactions. This study introduces a taxonomy of interactions to be used in association with the Functional Resonance Analysis Method (FRAM). The taxonomy has nine criteria: nature of agents, output nature, levelling, waiting time, distance, degree of coupling, visibility, safety and/or security hazards, and parallel replications. For each criterion, two descriptors are proposed: what the interaction looks like; and - when applicable - the variability level of the interaction. The use of the taxonomy is presented for three systems with clearly distinct complexity characteristics: cash withdrawal from an ATM, teaching a university course, and manufacturing operations. These case studies indicate the usefulness of the taxonomy for the identification of leverage points in work system design. They also show the value of modelling the variability of the interactions in FRAM models, in addition to the traditional modelling of the variability of the outputs of functions. Implications of the taxonomy for resilience engineering are discussed.

中文翻译：

---

社会技术系统中的交互作用的分类法：功能视角。

尽管交互作用的建模长期以来一直是社会技术系统理论的核心，并且是理解弹性的关键，但仍缺乏交互作用的整体分类法。本研究介绍了与功能共振分析方法（FRAM）结合使用的相互作用分类法。该分类法具有九个标准：代理的性质，输出的性质，调平，等待时间，距离，耦合程度，可见性，安全和/或安全隐患以及并行复制。对于每个标准，提出了两个描述符：交互作用是什么样的；交互作用是什么？以及（如果适用）互动的可变性级别。三种类型的分类系统的使用均具有明显不同的复杂性特征：从自动柜员机提取现金，教授大学课程，和制造业务。这些案例研究表明，分类法对于确定工作系统设计中的杠杆点很有用。除了对函数输出的可变性进行传统建模之外，它们还显示了在FRAM模型中对交互作用的可变性进行建模的价值。讨论了分类学对弹性工程的影响。