This paper aims to elevate essential worker accounts of the introduction of AI technology amid the COVID-19 pandemic. Drawing from a mix of ethnographic observations, interviews, and participatory design encounters with frontline staf, we examine the experiences of workers in a waste management facility in the United States newly tasked with overseeing autonomous foor cleaning robots. To complement and extend managerial and engineering descriptions emphasizing the functionality and performance of these devices, we used recuperative approaches to re-center the sociomaterial realities of workers on-the-ground. For example, workers reported concerns on the safety of the devices in congested areas and a need for more comprehensive training across all levels of the organization. This research seeks to expand the discourse on ethical AI by situating essential workers as a key source in developing best practices for deploying new technologies and evaluating pilot projects.

This paper examines the rapid introduction of AI and automation technologies within essential industries amid the COVID-19 pandemic. Drawing on participant observation and interviews within two sites of waste labor in the United States, we consider the substantial effort performed by frontline workers who smooth the relationship between robotics and their social and material environment. Over the course of the research, we found workers engaged in continuous acts of calibration, troubleshooting, and repair required to support AI technologies over time. In interrogating these sites, we develop the concept of "patchwork": human labor that occurs in the space between what AI purports to do and what it actually accomplishes. We argue that it is necessary to consider the often-undervalued frontline work that makes up for AI's shortcomings during implementation, particularly as CSCW increasingly turns to discussions of Human-AI collaboration.