Access Controls for IIOT Work Systems – Biometric Mechanisms

Abstract

This research aims to address security challenges in the IIoT (Industrial Internet of Things) work environment. The increasing digitalisation of industrial workplaces through the Industrial Internet of Things (IIoT) has transformed the way organisations manage physical access to critical infrastructure, manufacturing plants, and hazardous operational environments. Despite these advances, existing access control mechanisms, including passwords, RFID cards, tokens, PINs, and unimodal biometric systems, remain vulnerable to credential theft, spoofing attacks, environmental interference, and operational constraints. Industrial environments further complicate authentication due to the use of personal protective equipment (PPE), variable lighting, dust, machinery noise, worker movement, and harsh working conditions, which significantly reduce biometric recognition accuracy. Existing authentication solutions, therefore, fail to provide the level of security, usability, and resilience required for reliable contactless physical access control in modern IIoT environments.

This research addresses these limitations by designing, developing, and evaluating a contactless multimodal biometric fusion framework for secure physical access to industrial work sites. The proposed framework integrates facial, iris, ear, gait, palmprint, fingerprint, and voice biometrics within a deep learning-based fusion architecture capable of improving authentication robustness under dynamic operational conditions. To strengthen biometric template security, the framework incorporates biometric watermarking to protect identities and ensure data integrity. The research delivers three complementary Design Science Research artefacts: Artefact 1, a conceptual framework derived from a systematic literature review that establishes the theoretical foundations of contactless multimodal biometric authentication; Artefact 2, an experimentally validated prototype implemented within the Second Life virtual environment to simulate realistic industrial access control scenarios; and Artefact 3, a refined adaptive authentication framework informed by industrial expert evaluation, integrating contextual intelligence and adaptive biometric fusion to improve authentication resilience in complex industrial environments.

The study adopted the Design Science Research Methodology (DSRM) to create three prototypes, to guide the iterative design, development, demonstration, and evaluation of the proposed solution. The research combined a systematic literature review, conceptual framework development, prototype implementation, controlled

experimentation, and expert evaluation. A virtual industrial environment developed within Second Life enabled safe, repeatable, and cost-effective simulation of industrial access control without disrupting operational facilities. The proposed artefacts were evaluated for authentication accuracy, precision, recall, computational performance, environmental robustness, and usability, in accordance with ISO 9241-11.

The findings demonstrate that multimodal biometric fusion substantially improves authentication performance compared with conventional unimodal biometric systems while reducing authentication errors and increasing resilience against spoofing and operational variability. Experimental results confirmed that environmental conditions, including PPE occlusion, worker movement, poor illumination, and industrial noise, remain major factors influencing biometric reliability, thereby validating the need for adaptive authentication mechanisms. Although multimodal fusion significantly enhanced recognition performance, the research further established that static fusion alone is insufficient under highly dynamic industrial conditions. Instead, adaptive contextual weighting and intelligent decision-making were shown to improve authentication robustness by responding to changing biometric quality and environmental constraints. Expert evaluation further confirmed the practical feasibility, scalability, and industrial relevance of the proposed framework for future IIoT deployments.

This research contributes a validated contactless biometric access control framework, a comprehensive experimental evaluation methodology, and three rigorously developed Design Science Research artefacts that collectively advance knowledge in secure IIoT authentication. The research demonstrates how contactless multimodal biometric fusion can provide secure, usable, and scalable physical access control for industrial work sites while addressing real-world operational challenges. Beyond its immediate contribution to industrial cybersecurity and access management, the research establishes a foundation for future investigations into explainable artificial intelligence, adaptive biometric fusion, continuous authentication, privacy-preserving biometric protection, federated learning, and deployment within real-world cyber-physical industrial systems. The proposed framework, therefore, offers both a significant academic contribution and a practical pathway towards next-generation secure contactless biometric authentication for Industry 4.0 and emerging Industry 5.0 environments.