Smart manufacturing
Intelligent manufacture (Intelligent manufacture) is relative to mass manufacture. For nearly a century, the production methods of specialized division of labor and mass manufacturing have been widely adopted by the global manufacturing industry. The advanced manufacturing countries in the West are facing low cost and mass production from developing countries, especially Asia Manufacturing challenges. In order to improve the competitiveness of the domestic manufacturing industry on the one hand, and on the other hand to solve the problems in mass production (Mass Production), they proposed the concepts of smart manufacturing and Industry 4.0. The core of intelligent manufacturing proposed by scholars in Western countries is the transformation of M mass production as mass customization. They look forward to being able to produce various customized products immediately according to customer needs. For example, dozens of different configurations of cars are launched at once. Or a laptop with various hardware configurations and pre-installed software customized for customers.
This shows that intelligent manufacturing is not simply "machine substitution"-adding robots to the production line to improve the automation capabilities of the manufacturing industry. The automobile production line is a typical example. A large number of manipulators and automatic machines have been used in the automobile production line. But the current automobile manufacturing production line is still designed for mass production. It is not suitable for mass customization requirements.
The purpose of intelligent manufacturing is to meet the requirements of flexible production, and the realization of flexible manufacturing is a modular production line. The modular production line is composed of multiple manufacturing modules, each of which can be more productive, process requirements, and flexible configuration.
As various controllers in the manufacturing module, a distributed system is formed. In many cases, they are heterogeneous computer networks. Some of them may be PLC devices, inux industrial computers, Windows PCs. Building a flexibly configured distributed system for a modular production line is no small challenge. Under the concept of Industry 4.0, they are called Cyber-Physical System (Cyber-Physical System), and CPS in manufacturing system is called CPS production line (Cyber-Physical Production system or CPPS). Building such an innovative CPS production line is the goal of intelligent manufacturing.
From the perspective of information technology, the difference between CPPS and traditional production line control systems is that CPPS needs to make high-level planning and scheduling according to the production plan, and dynamically allocate (or configure) resources. Tailor craftsmanship for specific production. Modify the parameters and processing procedures of the manufacturing unit.
Application of Container Technology and IEC61499 in Intelligent Manufacturing System
In a heterogeneous distributed intelligent manufacturing network, the situation becomes extremely complicated. Different equipment adopts different network protocols and control methods. If the traditional centralized control method is adopted, different protocols require various protocol conversion equipment. Equipment manufacturers often work independently. In the process of system integration, sometimes they are equipped to control a device. A PC has serious software "hardening". Not to mention the increase in cost, it also increases the complexity of the system. Even so, the central control program will become very large, the efficiency of research and development, maintenance and system update is low, and the cost of developing a stable operation system is high.
The solution can adopt the concept of software decoupling in IT technology, and adopt container and microservice technology. Convert various protocols and place control programs in different containers. They will not interfere with each other. It can be developed, deployed and upgraded independently. For example, in the machining industry, there are various CNC equipment on the production line. They may use different control protocols. FUNC is their own protocol or MT-Link protocol. Or domestic machine tools may use Modbus protocol. Similarly, the control methods of robot arms of different manufacturers are in different ways and are not compatible with each other. In the traditional control system, a lot of program libraries will be added to deal with the equipment of different manufacturers. Or use various gateway devices. If the container method is adopted, we use an edge server device to deploy protocol conversion programs of different manufacturers in the container.
Of course, placing various protocol conversions and controls in the container does not completely solve the compatibility problem. Easily use them to write and modify applications is the key. The microservices in these containers need to have a unified packaging and presentation method for control engineers. The IEC61499 standard is a good choice.
IEC61499 is a function block programming method for distributed industrial measurement and control systems. And adopted an event-driven synchronization method. We can encapsulate each microservice in the container into one or more IEC61499 functional blocks.
Figure-1 is a schematic diagram, the actual function block will be much more complicated.
From the above schematic diagram, we can see that each FUNC CNC is described as an IEC61499 function block. In fact, it is a digital abstraction of physical CNC. It connects to the physical CNC equipment through the FUNC CNC microservice in docker. Similarly, the ROBOT ARM function block is a digital model of the FUNC robotic arm, which is connected to the physical FUNC ARM through the ARM microservice in docker.
Control engineers mainly use CNC and ARM function blocks to write TASK function blocks (it can also be a sub Application).
Here we have omitted the function blocks of the human-computer interaction interface (HMI).
Software architecture inside edge server
Figure-2 Edge server software architecture
As can be seen from the above figure, this is a Client/Server architecture. The edge server is the server side. Processing information is issued, system monitoring and production scheduling are completed through client software. The Docker container image can be completed using the docker client program.
The IEC611499 application software written by the control engineer is executed when the IEC61499 is running.
Example: Intelligent CNC production cell
Here we take the intelligent manufacturing production line in the machinery industry as an example to describe the workflow and architecture of CPPS.
equipment
Suppose a mechanical processing plant is composed of multiple CNC machine tools. It takes about 4 processes to process a mechanical part. In order to move to the intelligent manufacturing method, the company builds a robotic arm with every 4 CNC plus one FUNC into a manufacturing unit. The FUNC robotic arm completes the loading and unloading of CNC machine tools.
In such a manufacturing unit, different controllers of different devices are involved. mainly includes
- Most of CNC controllers are CNC controllers provided by FUNC.
- Execution parts in CNC, including axis servo drive, PLC, etc.
- FUNC robotic arm controller.
- Parts tray, AGV handling trolley, or conveyor belt.
Process
1 According to the CAD information of the parts, the workshop process design engineer designs the CNC machining program. Usually it is a CNC program written by G code. Preparation of production plans and dispatch orders.
2 The production scheduling issues a work order, and the engineering department downloads the G code to the relevant CNC controller.
3 The craftsman must formulate the processing sequence of the 4 CNCs and related process information.
Information system architecture
Each manufacturing unit is equipped with an edge controller. The controller accepts G Code and processing information sent by high-level software, and controls the processing process of the production unit. The edge controller is connected to CNC, robotic arm, and client computer via Ethernet.
Figure-3 A CNC manufacturing unit is composed of 4 Harting CNCs and a FUNC robotic arm
Support services for edge servers
The edge server is based on the docker microservice architecture similar to Figure-2. The main support modules include:
-FUNC CNC control service
-FUNC ROBOT control service
-Human machine interaction service
-Restfull API service,
Responsible for issuing CAD files and processing orders to the internal database.
-Database service
-Processing order service
Read the processing order in the database.
Concluding remarks
Intelligent manufacturing software based on docker container technology is more suitable for agile development and rapid deployment and update than traditional centralized single control software. Reduce various protocol conversions and gateways. Unify distributed manufacturing system equipment into an open system based on the LInux platform. , Conducive to system expansion, maintenance and backup. IEC611499 function block programming is suitable as the encapsulation and presentation of microservices. Meet the secondary development of OT engineers on the production line.
The CNC manufacturing cell project mentioned in the article needs to be further deepened. Errors are inevitable at present.