Key Challenges and Emerging Technologies in Industrial IoT Architectures: A Review

Abstract

The Industrial Internet of Things (IIoT) is bringing evolution with remote monitoring, intelligent analytics, and control of industrial processes. A reference architecture provides the general layout information for the flexible integration of IIoT systems; however, as the industrial world is currently in its initial stage of adopting the full-stack development solutions with IIoT, some challenges need to be addressed. To cope with the rising challenges and provide the blueprint guidelines to develop and implement IIoT in real-time, researchers around the globe have proposed IIoT architectures based on different architectural layers and emerging technologies. In this paper, we first review and compare some widely accepted IIoT reference architectures and present a state-of-the-art review of conceptual and experimental IIoT architectures in literature. We highlight scalability, interoperability, security, privacy, reliability, and low latency as the main IIoT architectural requirements and compare how the current architectures address these challenges. We also highlight the role of emerging technologies in current IIoT architectures to address these requirements and present the literature gap for future research work to address the challenges.

The Impact of Edge Computing on the Industrial Internet of Things

“…Other cloud computing limitations include the location of the centralised servers, as some countries may have laws prohibiting the storage of critical data in other countries (Basir et al, 2019). Computing and storage resources in a cloud computing setup may be exhausted as all data collected by IIoT sensors is transferred to a centralised server for storage and processing (Mirani et al, 2022). To offload the burden on the servers, processes must be in place to determine the criticality of the data before processing.…”

“…A fog computing network comprises nodes closer to the IIoT environment, with each node acting as a lightweight version of a cloud computing server (Caiza et al, 2020). This is done to offload the processing burden and shorten the time it takes to process and analyse the collected data (Mirani et al, 2022). In the management of IIoT environments, fog computing outperforms traditional cloud computing.…”

“…Edge computing provides an advantage over cloud and fog computing by allowing data processing at the edge, reducing the resources required for cloud computing by sending only the required data to the cloud for further processing (Mirani et al, 2022). Edge computing reduces data transfer latency, allowing for quick responses to real-time events.…”

“…However, delays in data transfer to centralised servers made it difficult for IIoT devices to respond to real-time events promptly (Basir et al, 2019). Edge computing was proposed as a solution for reducing data transfer latency by performing computations on the network's edge, which is close to the data source (Mirani et al, 2022). Only the necessary data is transferred to the centralised server for further processing in an edge computing environment, reducing the latency associated with data transfer (Qiu et al, 2020).…”

“…Since critical infrastructure sectors cannot tolerate delays in responding to real-time events, traditional IIoT setups that rely on a central server for storage and processing are insufficient (Qiu et al, 2022). By processing the data closer to the source, edge computing lowers the latency associated with data transfer to a centralised server (Mirani et al, 2022). However, edge computing is not without its critical flaws.…”