Spatial-temporal traceability and cyber-physical visibility for operations management

Abstract

Operations in the Industry 4.0 era are exposed to the anticipation of higher efficiency, agility, synchronicity and cost-effectiveness to meet the growing demand. Thus, it catalyzes the paradigm of smart manufacturing and intelligent logistics that aim to actualize informatization, digitalization, and intelligence for operations management. Specifically, it has been unveiled that enhancing spatial-temporal traceability (STT) and cyber-physical visibility (CPV) paves the way for operational improvement, as most operations should be conducted in the right place at the right time, and making information readable and operable could boost convenient handling and responsive decision-making. Accordingly, given that most operations take place indoors, this study attempts to integrate indoor positioning system, Internet of Things (IoT) and digital twin (DT) technology to acquire spatial-temporal information and make it visible in a timely and seamless manner. Then, it can assist operations conduction and management in effect. To begin with, a general framework of industrial indoor positioning system (IIPS) is proposed, which is dedicated to indoor localization in the industrial environment where interferences and noises proliferate greatly. In particular, this system comprises perception, interoperation, synchronization, and application (PISA) layers, involving a variety of advanced technologies. A novel indoor positioning approach is developed, namely genetic indoor tracking algorithm (GITA), which is enlightened by biological science. Amid, the fingerprinting technique based on Bluetooth Low Energy (BLE) technology is devoted to daily positioning, and deep neural networks (DNN) are designed for location estimation to pursue a better localization performance. Besides, Ultra-wideband (UWB) technology is also applied but just to sample labelling in supervised learning in order to alleviate the effort involved in system deployment and reduce the impact on regular operations. Moreover, a feature selection method and self-adapting mechanism are devised to reinforce and maintain long-term location accuracy and effectiveness. Two specific real-life scenarios are involved to validate the practicality and rationality of the proposed system and methods, namely factory logistics and cold chain logistics. Overall, several primary contributions of this thesis can be summarized. First, the IIPS with a PISA structure is proposed to enable spatial-temporal traceability and cyber-physical visibility so as to promote operational efficiency. The IoT, DT, and DNN techniques are utilized to fortify the whole system’s interconnectivity, synchronicity, and intelligence. Second, the GITA algorithm, inspired by genetics, is designed to proactively track objects in a real-time fashion using BLE technology, with the help of UWB for sample labelling. This algorithm is proved more stable, cost-effective, energy-efficient and easy-to-deploy. Third, the developed system and method have been applied to finished goods logistics in practice for illustration. The results show that the time spent on searching for products in the workshop is considerably reduced, and order emergency well controlled. Fourth, they also benefit cold chain logistics regarding occupational safety management in cold storage, paperless operations for shipment, and product quality monitoring in transit. In both two real-life cases, the operational efficiency has been noticeably enhanced. The total solution framework holds the promise of helping practitioners and scholars conduct reproductions to meet similar needs in reality.