The overall system architecture of smart manufacturing in the pharmaceutical industry can be divided into four layers: the equipment layer, control layer, business management layer, and operation management layer. In this article, we will introduce the intelligent systems in the equipment layer and control layer.

Equipment Layer

The equipment layer mainly includes three parts: production process systems, utility engineering systems, and automation control systems.

Production Process Systems

During the equipment selection process, attention should be given to the preparation of User Requirement Specifications (URS), especially emphasizing the requirements for technical evaluation. Attention should be paid to data standards, interface standards, and communication protocols for programmable controllers (PLC), distributed control systems (DCS), and human-machine interfaces (HMI) to ensure data interoperability between equipment and systems.

In the process of transforming traditional production workshops, efforts should be made to minimize manual data or information input by intelligently collecting data through sensors, smart instruments, and other means. For companies with the capability, advanced process equipment such as industrial robots, automated guided vehicles (AGV), vision recognition, and online testing can be considered.

Utility Engineering Systems

Companies should consider establishing a cluster-based utility engineering self-control system, covering aspects such as refrigeration units, cleanroom HVAC systems, water systems, electrical systems, and steam systems. This system comprehensively covers and manages clean environment control, water, electricity, gas, and steam supply and scheduling in the pharmaceutical industry to reduce energy consumption and improve production efficiency.

The construction of a cluster control system can be achieved through process control system technologies such as distributed control systems (DCS) and communication means such as industrial Ethernet to achieve interconnection of utility equipment. Standard communication interfaces should be reserved to provide data interfaces for data collection and supervisory control and data acquisition systems (SCADA).

Control Layer

Supervisory Control and Data Acquisition (SCADA) Systems

SCADA systems are primarily used for data integration to address the issue of data isolation. SCADA systems provide data to other information systems, transforming the previously decentralized monitoring mode into centralized control of various systems. Additionally, SCADA systems can promptly alert deviations and faults that occur during production, changing post-event management to real-time online management.

SCADA systems are typically designed for the entire factory using industrial configuration platform software. Considering the future requirements for data collection volume and collection cycle performance, a functional division architecture can be adopted, with data collection and data storage configured on separate servers with dual-machine hot backup. Given the pharmaceutical industry's requirements for data integrity, reliability, and real-time performance, industrial real-time databases are generally used as the foundational storage pool for enterprise information systems, providing basic data for other information systems.

SCADA systems generally include real-time monitoring of production processes, recording of abnormal alarms, data processing and analysis, real-time data management and archiving, and access control. Data processing and analysis are the core of SCADA systems, consisting of functions such as data collection, real-time databases, event alarms, historical trend databases, and historical data/event alarm archiving. Furthermore, audit tracing functionality should be configured to ensure traceability of operations.

Automation Control Systems

The production process of pharmaceutical companies is typically divided into two stages: the production of raw materials and the production of finished pharmaceuticals. The process control of raw materials is usually completed using large distributed control systems (DCS), while in the production of finished pharmaceuticals, the equipment used in various process steps is typically controlled by embedded PLC systems. These PLC systems often come from different manufacturers and operate independently. With the continuous penetration and spread of the Internet and new-generation information technology, the previously relatively closed operating environment of industrial control systems has gradually been broken, leading to greater openness and interconnectedness. Automation control systems collect field control data and information by establishing interfaces with DCS or SCADA systems and interface with Manufacturing Execution Systems (MES) to achieve information flow.

Key components of the automation control system, such as power supplies, controllers, and networks, can be configured redundantly in a 1:1 ratio. They are equipped with UPS power supplies and AC power switchers to ensure the safety and stability of power supply, communication, and CPU. Time synchronization is achieved for all computers within the local area network. In accordance with GMP requirements, each authorized user has a unique identification code, password, and electronic signature. The system allows for electronic signature and audit tracing of operations.

The automation control system can use flexible batch formulations to control the production process. Based on the ISA-S88 standard, it provides consistent standards and terminology, defines physical models, procedures, and formulations, controls batch process production, manages the entire production manufacturing process from raw material input to product delivery, and allows for efficient management of production parameters for different batches through interaction with MES. The automation control system feeds back data such as time, status, and process parameters to MES, completing electronic batch records.

The automation control system can guarantee special environmental requirements. Following the GMP management concept, an intelligent control system for production environments is constructed to create a "one-stop" cluster-based utility engineering control system. This system ensures the supply and control of energy resources such as electricity, water, and gas and maintains a clean environment. Specific components include clean air conditioning and purification systems, cleanroom environmental monitoring systems, personnel access control systems, video surveillance systems, intelligent environmental protection treatment systems, and more. Additionally, an integrated Safety Instrumented System (SIS) can be established to ensure safe production. Based on a distributed control system, advanced, applicable, and effective professional calculation methods are used to improve system reliability.

The automation control system can also introduce Process Analytical Technology (PAT) to collect real-time quality data from the production process, enabling product quality control. The system utilizes online process analysis instruments to perform real-time measurements of critical quality parameters and performance characteristics for raw materials, materials in process, and key quality attributes during the manufacturing process. This information is used to design, analyze, and control the production and processing processes and accurately determine the quality of intermediate and final products.

Furthermore, Machine Vision Recognition Systems and industrial robots provide new possibilities for automation control systems in pharmaceutical companies. Machine Vision Recognition Systems use machines to replace human eyes for measurement and judgment, simulating human visual functions by capturing images with computers to extend human vision. Machine Vision is used to observe the production site, determine results, and classify them. Industrial robots, in combination with robotic arms and automated transport technology, modularize and automate the production process, improving production efficiency and enhancing workshop working conditions.

To facilitate the deep mining and organization of data in the later stages, a Plant Information System (PI) application platform should be added to the automation control system as a large-scale real-time database and historical database. It is used for the automatic collection, storage, and monitoring of factory data, providing clear and accurate operational screens. Users can view the current production situation and review past production situations, offering efficient factory information for end-users and application software developers to improve processes, implement comprehensive quality management, and execute preventive maintenance, among other tasks. (Excerpt from the "Chinese Pharmaceutical Industry Smart Manufacturing White Paper (2020 Edition)," jointly compiled by the Ministry of Industry and Information Technology Industry Development Promotion Center and the China Association of Pharmaceutical Commerce)