Comparing Static and Dynamic Human-Robot Task Sharing - Experiment Software
Year: 2026
Type: software
Abstract: In human-robot collaboration (HRC), task scheduling determines how assembly steps are assigned to agents. These schedules can be static, generated prior to execution, or dynamic, adapting to events in real-time. Despite the proliferation of novel scheduling algorithms, the robotics field lacks standardized, reproducible methods for evaluating how human collaborators experience these competing policies. Existing user studies suffer from heterogeneous hardware setups, proprietary software stacks, and unshared study materials. We address this reproducibility gap by introducing a complete, open-source platform designed for HRC scheduling experiments. The provided artifact includes the hardware setup, graphical user interface, study documents, and a modular software architecture with a dedicated scheduler interface to test custom algorithms. Utilizing this platform, we conducted a within-subjects user study (n = 22) to compare static and dynamic task sharing during an assembly scenario with experimentally induced human absences. Our results demonstrate that dynamic scheduling significantly reduces the user's perceived workload and increases perceived team fluency, without raising mental demand.
Institution: University of Bayreuth
Language: English
Subfield: Control and Systems Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7203750187