Environmental Intelligent Perception in the Industrial Internet of Things: A Case Study Analysis of a Multicrane Visual Sorting System

Future environmental perception (EP) in the Industrial Internet of Things (IIoT) is expected to interconnect massive sensing objects, merge various intelligent technologies, and efficiently control devices to adapt to a complex environment. It will optimize overall production value, including improving product quality, promoting multisensor collaboration, and decreasing energy consumption and costs. Current EP perceives the industrial environmental information through industrial sensors that are controlled by traditional programmable logic controllers (PLCs), but the sensors are not capable of collaboration and intercommunication. Some sensors are difficult to deploy in the terminal and can easily be corrupted in a harsh environment. In addition, the current EP lacks intelligent analysis of massive sensing data. Therefore, it is necessary to use the latest emerging techniques including knowledge graph (KG), artificial intelligence (AI) algorithms, and cloud control, to create environmental intelligent perception (EIP) in the industry. In this article, an overview of challenges for EP in a complex industrial environment is provided, potential schemes for designing EIP architectures are given, and potential practical application scenarios are discussed. Finally, a case study analysis of a multicrane visual sorting system is presented in the experiment, in which two mechanical arms operate cooperatively with the help of cloud PLCs and visual recognition algorithm.