A Cloud-Based Intelligent Nursing Care System with a Soft Robotic Hand for Autonomous Hygiene

This paper presents the development of a cloud-based intelligent nursing care System (CBINCS) integrated with a vision-guided soft robotic hand to provide safe, hygienic, and dignified post-defecation care for elderly. The CBINCS is developed using an integrated mechanical, control, and machine-vision framework. It consists of a robot arm, a Soft Robotic Hand (SR-Hand), a depth camera, and Cloud Server (ACS). The SR-Hand combines ultrasonic vibration cleaning, water spray, and tendon-driven actuation to enable adaptive perianal hygiene. The depth camera acquires synchronized RGB-D images of the perianal region to detect fecal residues, which are transmitted via a Wide Area Network to the ACS. The cloud server performs fecal detection and 3D localization, and the results are sent back to the robotic system to guide precise cleaning actions. Experimental validation was conducted using an anatomical plastic rubber model and artificial feces. The results confirmed a localization repeatability of 0.52 mm RMS within the ±15 mm tolerance of the SR-Hand. Cleaning success rates were 98.8% in open conditions, 93.5% in cluttered conditions, and 82.3% in overlapping feces conditions. Average cycle times were 8 s for data acquisition and 6 s per cleaning cycle. Cloud latency measurements ranged from 8.0 to 15.2 s depending on network conditions, confirming the proposed latency model. Cost analysis revealed that the cleaning robot, designed using low-cost processors, can leverage cloud resources, avoiding expensive onboard GPUs and enabling scalable multi-robot deployment .