Search

Smart factory adopts cyber-physical technologies integrating independent discrete systems into a context-sensitive manufacturing environment to optimize manufacturing processes using decentralized information and real-time communication. This paper presents our work in the realization of a smart factory, which comprises of a four-layer hierarchical architecture, i.e. connection infrastructure, data acquisition, data processing and smart applications. In the connection infrastructure layer, all shopfloor machines are connected through diverse protocols, IoT sensors, PLC interfaces and DNC connectors. A centralized IoT gateway supports such a scalable and adaptable connection and ensures a reliable communication among all heterogeneous manufacturing systems. In the data acquisition layer, the real-time machine and job data are acquired from shopfloor systems. Machine data indicates machines’ working condition and job data reveals the production information. The data processing layer comprises of three modules, i.e. shopfloor monitoring, data visualization and data analytics, which monitor and visualize shopfloor activities and analyze the semantics of various data using AI-based TPM engines providing the scientific indicators for next-step decisions. The smart application layer provides with several decision-making and remote control functions for shopfloor productivity and efficiency, such as predictive maintenance, shopfloor management, machine & job optimization and digital twin. The smart factory system has been implemented in the manufacturing shopfloor at Nanyang Polytechnic. The results and validation show that the system can simultaneously collect and analyze the manufacturing data from shopfloor systems, and further communicate with and control the shopfloor systems with decision-support functions. The overall shopfloor efficiency and flexibility can be significantly improved towards a smart factory of Industry 4.0.

The ability to instantly and successfully interact among technical engineers, management team and machines is vital to the productivity and efficiency of manufacturing shopfloor. Chat messenger with smart mobile has obvious advantages for shopfloor management in group communication, instant notification and remote control. This paper presents a natural-language chat & control HMI for manufacturing shopfloor. The system successfully realizes a smooth two-way communication between users and shopfloor machines. Users can access just-in-time information, receive instant notifications, and remotely control shopfloor machines. All relevant parties can communicate over shopfloor matters in a chat group. The system comprises of four core modules, i.e. Chat Messenger, Chat Service Engine, Control & Communication Engine and Local Command Service. Chat Messenger provides chatting user interface and group management dealing with end-user’s enquiries and notifications via natural language. Chat Service Engine and Control & Communication Engine are two cloud-based service modules, which process questionnaire logic and transmit relevant commands and data bi-directionally between Chat Messenger and Local Command Service. Local Command Service is a local service terminal which implements interfaces and protocols directly interacting with shopfloor machines. It processes the requests and commands from Chat Messenger to shopfloor machines. It also checks real-time machine anomalies and automatically generates corresponding notifications. The system has been implemented in the advanced manufacturing shopfloor at Nanyang Polytechnic. The results and validation show the improvement of manufacturing shopfloor efficiency.

In this highly globalised manufacturing ecosystem, product design and verification activities, production and inspection processes, and technical support services are spread across global supply chains and customer networks. Therefore, a platform for global teams to collaborate with each other in real-time to perform complex tasks is highly desirable. This work investigates the design and development of a remote real-time collaboration platform by using human motion capture technology powered by infrared light based depth imaging sensors borrowed from the gaming industry. The unique functionality of the proposed platform is the sharing of physical contexts during a collaboration session by not only exchanging human actions but also the effects of those actions on the task environment. This enables teams to remotely work on a common task problem at the same time and also get immediate feedback from each other which is vital for collaborative design, inspection and verifications tasks in the factories of the future.