The study aims to explore and construct an interaction scenario model for in-vehicle robots in intelligent cockpits, to offer both a theoretical framework and practical guidance for interaction design. The grounded theory was employed with a three-level coding approach to systematically analyze 13 publicly available product samples and 36 in-depth interview cases. Based on the coding results, a layered and progressive functional scenario model was developed. The open coding process yielded 39 categories, which were consolidated through axial coding into 15 major categories and further synthesized into six overarching functional scenarios: driving safety, intelligent services, direct settings, voice interaction, emotional services, and human-subjective interaction. The theoretical model derived from selective coding reveals a three-tiered structure of "demand-driven—interaction realization—experience enhancement". Findings suggest that interaction scenarios of in-vehicle robots form a systematic framework encompassing safety, functionality, and emotionality. Research shows that safety assurance serves as the fundamental prerequisite, functional realization acts as the central support, and emotional experience emerges as the value extension. This model provides a structured theoretical basis for interaction design in intelligent cockpits and offers practical guidance for the optimization and application of in-vehicle robot design in the future.