← CHI 2026 map

CHI '26 · Honorable mention · full-paper review · confidence medium-high

Thing Ethnography in a Factory: Exploring Emergent and Dynamic Relations of Cobots and Workers

Hyungjun Cho , Jiyeon Amy Seo , Yongjae Sohn , Hee Rin Lee

DOI PDF Program page Page JSON

This is a strong qualitative CHI paper because it does not merely describe cobots in a factory; it re-frames what a cobot is in practice. The main contribution is conceptual and empirical: a relational account of three enacted roles, backed by a focused field study, with clear limits on transferability.

Axes Lens

Rare contribution shape, typical evidence profile. The point here is not a score. It is to show what kind of claim the paper makes, and whether the evidence pattern is unusual or baseline in this 268 -review set.

Review Summary

This paper stands out because it pushes against a very common industrial-HCI assumption: that cobots are best understood as preconfigured collaborators whose value lies in autonomy and efficiency. The authors instead use a robot-centered thing ethnography to show that cobot roles are not stable properties of the machine but emergent outcomes of relations among the cobot, CNC machines, workers, and the factory owner. That move is both the paper’s conceptual contribution and its empirical payoff. The three roles they identify—frontline operator, care receiver, and scapegoating manager—are memorable because they make visible forms of work and dependency that are usually backgrounded in human-centered accounts. Methodologically, the study is credible for a CHI qualitative paper: it is grounded in an 18-day observation period and a large corpus of automatically captured clips, which supports the descriptive claims about recurring patterns. At the same time, the paper is appropriately bounded. The authors explicitly acknowledge that they lacked worker interviews and that access constraints limited in-person observation, so the interpretation is not triangulated in the way a mixed-methods field study might be. That limitation matters, but it does not undermine the core contribution; rather, it clarifies that the paper is strongest as a situated, theory-informed empirical re-description of cobot work relations in one factory context. Overall, this is a solid honorable-mention-level contribution because it offers a sharp reframing, a well-matched method, and a clear set of boundary conditions without overreaching beyond the evidence.

What Changed

Canon before

Cobots in industrial HCI are commonly framed as autonomous collaborators or productivity tools with stable roles, often analyzed through human-centered accounts of work and coordination.

Departure from common sense

The paper challenges the default assumption that cobots have fixed, pre-defined collaborative roles. Instead, it argues that their roles are produced dynamically through relations with machines, workers, and the owner, which is a notable shift from treating the robot as a stable autonomous agent.

Actual novelty

The paper’s novelty is the combination of robot-centered thing ethnography with an intra-action lens to surface three context-dependent cobot roles—frontline operator, care receiver, and scapegoating manager—showing that cobot agency is enacted relationally rather than assigned in advance.

Evidence

The evidence comes from an 18-day robot-centered thing ethnography in a small-scale factory, with automatically recorded short video clips and analysis that yielded three relational role configurations. The paper also explicitly notes a methodological limitation: no worker interviews, and access constraints reduced in-person observation, leaving interpretation grounded primarily in video data.

Limits

Method limits

The study is bounded by a single small-scale factory setting and relies heavily on video clips rather than direct participant interviews. The authors note that the absence of worker interviews limits triangulation and may leave some interpretations underexplored.

Deployment limits

The findings are most directly applicable to similar industrial settings where cobots, workers, and surrounding machinery interact continuously. The paper does not establish that the three roles will transfer unchanged across factories, production lines, or organizational cultures.

Boundary conditions

The analysis is conditioned by a small-scale factory, an 18-day observation window, and event-triggered video capture. The relational roles identified may depend on local maintenance practices, machine configurations, and the specific social organization of work.

Manual check: false

Position in field

This paper contributes to CHI’s growing interest in posthuman, relational, and ethnographic accounts of workplace technology by shifting attention from human-centered collaboration narratives to the enacted roles of cobots within a production ecology.

Abstract

Collaborative robots (cobots) are often portrayed as transformative technologies that promise efficiency and productivity in industrial workplaces, with their roles frequently pre-defined as autonomous collaborators to human workers. This paper presents a different picture based on an 18-day robot-centered thing ethnography conducted in a small-scale factory. In practice, the roles of cobots were dynamically constituted through intra-actions among CNC machines, human workers, and the factory owner. Our analysis identifies three roles: (1) frontline operators carrying out repetitive tasks, (2) care receivers dependent on continuous maintenance and cleaning, and (3) scapegoating managers whose stoppages summoned human intervention and mediation. These relational configurations foreground dynamics that remain backgrounded in human-centered accounts and offer a deeper understanding of how cobots are enacted within production environments. We argue that understanding cobot integration and advancing practical design discourse requires moving beyond assumptions of collaboration and autonomy, recognizing cobots as entities continuously redefined through their relations with environments and other agents.