Human–Robot Interaction (HRI) research often relies on controlled laboratory studies to ensure rigour and repeatability. However, the recent rapid increase in robot deployments across diverse sectors has revealed that findings from controlled studies often fail to transfer to real-world performance. Outside the lab, human behavior is shaped by interruptions, adaptation, partial attention, long-term engagement, and social norms. Environmental uncertainty, sensor noise, and system limitations further complicate the reliable monitoring of human behavior, resulting in unanticipated interactions that are difficult to reproduce in controlled settings, sometimes putting human safety at risk. Human data collected in the wild is extra messy, sparse, heterogeneous, and ethically constrained, yet it reflects the conditions under which HRI systems must ultimately operate. However, learning from real human behavior introduces substantial methodological challenges, including limited experimental control, noisy and incomplete data, small and diverse participant samples, high deployment costs, and ethical, privacy, data-safety and data-ownership constraints. Despite these challenges, embracing in-the-wild HRI is essential for developing robust, adaptive, and deployable robotic systems.

Organized by researchers and practitioners studying a range of human behavior monitoring methodologies and applications in different HRI domains, this workshop provides a structured forum to critically examine methods for observing, modeling, and interpreting human behavior in real-world HRI over extended timescales. Through multidisciplinary perspectives, it encourages participants to rethink analytic pipelines, evaluation criteria, and assumptions underlying behavioral models, and to openly discuss in-progress ideas, lessons learned, failures, surprises and methodological tensions often overlooked in traditional venues. The workshop targets HRI researchers, roboticists, designers, and practitioners working in domains such as manufacturing, healthcare, education, public-space robotics, and disaster response, with contributors from academia and industry sharing practical insights from real-world deployments.

Banner Images Used From: https://techcrunch.com/2022/01/13/serve-robotics-new-autonomous-sidewalk-delivery-robots-dont-require-human-assist/, https://www.researchgate.net/figure/Collaborative-robot-working-with-a-human-to-assemble-an-engine-31_fig1_359492378, https://link.springer.com/article/10.1186/s12913-024-10857-9, CSIRO Robotics