Patient agitation, unplanned device removal, and restraint-related complications remain persistent safety challenges across hospitals, and long-term care facilities. Conventional restraint practices rely heavily on intermittent observation and subjective judgment, leading to delayed detection of escape attempts, pressure injuries, and physiological deterioration. Despite advances in digital health, few solutions integrate nurses’ clinical insights, physiological monitoring, behavioral assessments, and real-time decision support into a unified system.To develop an AI-enabled Multimodal Restraint Monitoring and Clinical Decision Support System (AI-MRM-CDSS) that unifies multimodal data to enable timely care and informed clinical decisions.Built on a patented multimodal monitoring framework (Patent Application No. 114132XXX), an interdisciplinary team from nursing, engineering, and informatics designed AI-MRM-CDSS with three core components: (a)Clinical Input Interface that consolidates agitation assessments (RASS, GCS, CAM-ICU), pain scores, tubing status, medication information, and scheduled release intervals; (b)Embedded Sensor Modules within restraint bands that continuously capture heart rate, oxygen saturation, and pull-force patterns, which are wirelessly transmitted to the system's edge-computing unit for processing; (c)Edge-AI Analytics Engine providing real-time escape-risk predictions, restraint adjustment recommendations, tension-misuse alerts, medication-related insights, and automated reminders to support routine release and pressure injury prevention.The system architecture, interface, and multimodal sensing framework have been completed. Predictive algorithms are expected to identify escape attempts, trigger real-time alerts, and provide personalized recommendations on medication dosing and restraint use. Automated two-hour reminders will support timely repositioning and help prevent pressure injuries. Upcoming work includes algorithm refinement, simulation testing, and usability evaluations to assess predictive accuracy, alert relevance, workflow integration, and clinician acceptance.Integrating nursing expertise ensures that AI-generated recommendations align with actual clinical demands. By combining physiological signals, behavioral assessments, medication information, and tubing status, AI-MRM-CDSS addresses a long-standing gap in restraint safety by providing continuous monitoring, timely alerts, and actionable decision support to enhance patient safety.AI-MRM-CDSS shows potential to enhance restraint safety and clinical decision-making. Its multimodal and scalable design allows application not only in hospitals but also in long-term care facilities. Planned testing and validation will further determine its utility across diverse care settings. Keywords: Artificial Intelligence, Decision Support System, Patient Safety, Physical Restraint