Increasing use of embodied AI and autonomous medical devices is transforming smart hospital IIoT/IIoMT into highly connected cyber-physical ecosystems. Unlike conventional healthcare authorization systems, smart hospitals require trustworthy governance and collaborative embodied agents coordinating communication, sensing, data access, and physical actuation under strict safety, accountability, and privacy requirements. However, available techniques heavily rely on centralized decision-making and static access control policies that limit the abilities of those approaches to address privacy, safety, dynamic trust, and accountability requirements. This article proposes a trust-adaptive explainable consensus and permissioned governance framework for collaborative embodied AI in smart hospital IIoT/IIoMT networks. The proposed framework incorporates specialized governance agents responsible for policy compliance, preserving privacy, clinical validation of health conditions, risk assessment, and safety evaluation. A trust-aware explainable consensus approach aggregates the agents’ recommendations for supporting accountable and transparent decision-making, while a human oversight module handles high-risk and uncertain situations. For ensuring governance integrity and auditability, decision metadata and trust records are stored on a permissioned blockchain. An adaptive trust management engine evaluates the reliability of the participating devices and agents based on governance and operational behavior outcomes: supported by experimental evaluation and validation.
The proliferation of the Internet of Things (IoT) applications has necessitated the inclusion of sensing functions in network architectures. To this end, integrated sensing and communication (ISAC) is emerging as a key use-case for sixth-generation (6G) networks. ISAC enables efficient utilization of shared resources for joint communication and sensing functions. However, accommodating massive networks with joint sensing and communication functions intensifies spectral congestion. Furthermore, this dual functionality creates a higher energy footprint which exacerbates sustainability concerns. In this regard, backscatter technology is emerging as a promising technology for spectrum-intelligent ISAC. Since backscatter tags modulate existing radio-frequency (RF) signals, they achieve spectrum reuse without introducing additional spectral interference to coexisting systems, thus, enabling low-power, sustainable operation. Moreover, their distributed reflections can enable distributed spectrum sensing that supports adaptive spectrum management in spectrum-constrained ISAC systems. Motivated by the above, this magazine article provides a comprehensive overview of the foundations, architectures, and configurations for backscatter-enabled ISAC (B-ISAC) systems. To further enhance feasibility, we conduct a deep reinforcement learning (DRL)-based case study for a novel fluid-antenna (FA)-aided B-ISAC system. We also compare the proposed scheme to a state-of-the-art (SoTA) system. Additionally, we articulate the challenges and future directions associated with these systems for underpinning sustainable and spectrum intelligent IoT networks in next-generation systems.
