Resilience is a capacity developed and deployed over time rather than an effective reaction to a single event (Hamel & Välikangas, 2003; Reinmoeller & Baardwijk, 2005). Such a capacity can be expressed at different levels of aggregation, spanning from individuals to entire communities (e.g. Van Der Vegt et al., 2015). The brilliant examination of the Mann Gulch disaster conducted by Karl Weick (1993) shows that even organizations – and systems - which are designed to face risky situations can result vulnerable when exposed to the unexpected, while experiencing “erosion of sense and structure” (p. 638). As Weick (1993) anticipates, organizational improvisation (Cunha et al., 1999; Hadida et al., 2015) provides the means for exploration of alternatives (e.g. Hummels et al., 2014) to face the unpredictable. As in many artistic and emergent practices (e.g., Furu, 2013; Vera & Crossan, 2003), flexible, extemporaneous, improvised action stimulates learning from the unexpected (e.g. Miner et al., 2001), that complement what routines cannot prescribe (e.g. Sutcliffe & Vogus, 2003). Hence organizational improvisation contributes to resilience as it turns resource constraints into opportunities, stimulating resourcefulness and imagination (Gibbert & Välikangas, 2004), leveraging upon organizational ingenuity and imaginative recombination of resources (Dolmans et al., 2014; Korhohen & Välikangas, 2014). The quest for resilience is particularly central in organizations and systems that are designed for high-reliability (e.g. Milosevic, Bass & Combs, 2015). Reality has proven, however, that even technologically sophisticated systems do fail. Failure occurs at the boundary of human and technology interactions in extreme environments. Among others, cases such as Mann Gulch or the AF447 flight (Oliver et al., 2017) show that surprises often creep through design, technology and coordination in unsuspecting manners. In line with previous work, we define surprises as events that happen unexpectedly or expected events that take unexpected shapes (Cunha et al., 2006). So, some question arises: How does the predictable become surprising for organizations designed to be resilient? How does such a design for resilience turn into vulnerability?