RABBIT is a robot-assisted bed bathing system that combines multimodal perception and dual (software and hardware) compliance to perform safe and comfortable physical human-robot interaction. The system uses RGB and thermal imaging to accurately segment dry, soapy, and wet skin regions, enabling effective execution of washing, rinsing, and drying tasks. It includes custom-designed motion primitives inspired by human caregiving techniques and a novel compliant end-effector called Scrubby, optimized for gentle and effective interactions. A user study with 12 participants, including one with severe mobility limitations, demonstrated the system's effectiveness and perceived comfort. The system uses a synthetic bathing perception (SBP) dataset for training segmentation models and includes a task-space compliance controller for software-level compliance. RABBIT aims to bridge the gap between advanced robotics and the nuanced requirements of caregiving by drawing inspiration from human caregivers and addressing the unique challenges of bed bathing. The system's contributions include the development of a robot-assisted bed bathing system with multimodal perception, motion primitives inspired by caregiving practices, and dual compliance. It also releases the SBP dataset and the open-source Scrubby end-effector. The system's performance was evaluated in a user study, showing high ratings for coverage, force application, and consistency. The results indicate that RABBIT can provide a comfortable and efficient bed bath, with participants reporting high satisfaction. The system's adaptability to different user conditions was demonstrated through interactions with a participant with Multiple Sclerosis. Future work includes improving the system's ability to handle residual soap and water, adapting to involuntary movements, and incorporating user preferences into the robotic system.RABBIT is a robot-assisted bed bathing system that combines multimodal perception and dual (software and hardware) compliance to perform safe and comfortable physical human-robot interaction. The system uses RGB and thermal imaging to accurately segment dry, soapy, and wet skin regions, enabling effective execution of washing, rinsing, and drying tasks. It includes custom-designed motion primitives inspired by human caregiving techniques and a novel compliant end-effector called Scrubby, optimized for gentle and effective interactions. A user study with 12 participants, including one with severe mobility limitations, demonstrated the system's effectiveness and perceived comfort. The system uses a synthetic bathing perception (SBP) dataset for training segmentation models and includes a task-space compliance controller for software-level compliance. RABBIT aims to bridge the gap between advanced robotics and the nuanced requirements of caregiving by drawing inspiration from human caregivers and addressing the unique challenges of bed bathing. The system's contributions include the development of a robot-assisted bed bathing system with multimodal perception, motion primitives inspired by caregiving practices, and dual compliance. It also releases the SBP dataset and the open-source Scrubby end-effector. The system's performance was evaluated in a user study, showing high ratings for coverage, force application, and consistency. The results indicate that RABBIT can provide a comfortable and efficient bed bath, with participants reporting high satisfaction. The system's adaptability to different user conditions was demonstrated through interactions with a participant with Multiple Sclerosis. Future work includes improving the system's ability to handle residual soap and water, adapting to involuntary movements, and incorporating user preferences into the robotic system.