Evaluation and implementation of zero-touch onboarding solutions for IIoT

Abstract

The Internet of Things (IoT) has provided numerous opportunities across various industries, and its technological advancements are rapidly progressing. The process of Device Onboarding refers to the task of initially registering a device onto an IoT cloud. Onboarding can be very challenging while doing a large-scale IoT deployment. A recent report by IoT analytics predicts that the number of IoT devices will increase twofold, with an expected total of Twenty-seven billion devices by 2025. As the number of IoT devices proliferates, establishing a streamlined process for onboarding multiple devices has become more imperative than ever. While doing this research, we first evaluated the state-of-the-art solutions for onboarding IoT edge devices. The methods were Fido (Fast Identity Online) Device Onboard (FDO) from the Linux Foundation, Keylime Solutions from the cloud-native computing foundation (CNCF), and OPCUA device onboard by the OPC foundation. The evaluation is based on 22 various factors. The assessment led us toward the best possible solution, i.e., FDO. to do zero-touch onboarding and register the device to an IoT device management service. In addition to addressing the existing inconsistencies in the device manufacturing process using FDO standards, this research has also implemented these standards in a novel way. Many zero-touch onboarding solutions today demand special software and hardware to be integrated into the IoT device during manufacturing. The proposed solution using FDO standards with a "Late Binding"feature is an open, cloud-agnostic solution that allows users to choose their preferred cloud provider during the initial power-up. It eliminates the need for Original Device Manufacturers to manufacture unique device Stock Keeping Units (SKUs) for each customer and cloud combination. To summarize, This research thoroughly evaluated and chose a suitable onboarding method and successfully demonstrated the implementation. We used the raspberry pi compute module 4 as an IoT edge device with a custom-embedded Linux OS. Developing the novel hawkBit onboarding modules on both the client and server sides has facilitated this implementation. The proposed software-based solution is capable of onboarding 43200 devices per day. Utilizing the FDO 1.1 standards, this approach has proven to be a highly effective solution for the zero-touch onboarding of a large number of IoT devices.