Factory Planning "SAIC Practice" based on digitalization and intelligent technology

Home > Sci-Tech

Factory Planning "SAIC Practice" based on digitalization and intelligent technology

2021-11-26 00:13:37 10 ℃

With the continuous improvement of labor costs, the market competition is increasing, and the automotive manufacturing industry has put forward higher requirements for TQCS (Time Time, Quality Quality, Cost Cost, Service Services). In order to address challenges, more and more automotive manufacturing companies use new technology-driven digitization and intelligence strategies to become an unstoppable trend.

(1) Traditional factory planning process

Based on market demand, automotive manufacturing companies developed product designated products, after planning plant in demand, investment, venue and other needs. The traditional factory planning process first conducts process planning based on product objects, and then performs patterns and optimization, and finally performs logistics, auxiliary areas and factory overall planning, and the process is shown in Figure 1.

Figure 1 Traditional factory planning process

1.1 Process Planning

Traditional process plan first needs to analyze the parts pattern, and organize the part characteristics to estimate the processing or assembly process. According to the engineer's personal experience, the primary version is arranged, and the size chain, tool selection, process arrangement, and processing program are optimized after repeated inspection. This process requires a relatively long period, and the technical planning quality is a big relationship with the personal experience of craft engineers.

1.2 beat analysis

The beat analysis is an important step in the process planning, which determines the assignment of the station operation content. In the process of beat analysis, standard work time analysis is required for handmade or semi-automatic stations. Using our standard working hours analysis tool such as MOD, MTM calculates the standard operation time involving manual operating stations to determine the personnel planning and work-time balancing wall. The standard working time is calculated to take a lot of time.

1.3 production line, logistics area, process aid room and factory plan

Combined with the principle of SLP (Systematic Layout Planning) system arrangement design and lean arrangement, the relevant production line, logistics area, and process auxiliary room plan to meet the needs of production. This process is based on two-dimensional CAD, there may be problems with equipment interference, human-machine risk, and abiners with abiners. After completing the production line, logistics, and the process auxiliary room plan, the factory design is launched, and the construction, structure, HVAC, electrical, power, water supply and drainage are complex, based on traditional two-dimensional design methods, may occur later Problems such as pipelines or facilities interference, misplaced dislocation, resulting in unnecessary rework. In the planning and design of the production line and logistics, the capacity assessment, logistics path, the number of tray, and the number of AGC are based on empirical calculation, logical complex, and the results may not be optimal.

(2) Digital chemical plant planning

Among the traditional factory planning processes, there is a problem of process document production, standard working hours computational fees, poor planning, high rework cost. Digital provides a solution for different linkages that may be involved in the above-mentioned traditional plant planning. With the continuous advancement of computer technology, as the main embodiment of digitalization - virtual modeling and simulation technology have achieved great development, and many links in factory planning are applied.

Figure 2 Process of digital chemical plant planning

In order to solve many problems in the traditional factory planning method, the digital chemical plant planning method came into being. Compared to the traditional way, digital tools are used in the plant plan, and the project cycle is expected to shorten the project cycle by 20% of the company's practical experience. The main process of factory digital planning is shown in Figure 2, first based on the product digital model, after obtaining the production line planning design, the production line planning design is performed, and the simulation verification of the production line is performed. The following will be set forth in the application of digital chemical plant planning.

2.1 Digital Process Planning

There are a variety of ways to achieve digital process planning. SAIC's General Motors use Siemens Product Lifecyclement, PLM, which is connected to product design and manufacturing process development. The process of digital process planning is from Teamcenter to introduce the product to the NX secondary development platform and PD / PS (Process Designer & Process Simulate), as shown in Figure 3, Figure 4, process planning, beats of machining and assembly Analysis, line balancing and machining process simulation, forming process flow chart, finally transferring the results back to Teamcenter for unified platform management, realizing the management of process planning and manufacturing information, thereby reducing the risk of process planning, improving the efficiency of process planning .

Figure 3NX secondary development machine processing digital development

Figure 4 TECNOMATIX Process Designer Assembly Process Digitization Development

2.2 Production line planning design

Based on the process flow chart and standard workers, combined with the principle of land, logistics regions and process aids in combination with SLP system arrangement and assembly process. Preliminary planning. In the layout design phase, three-dimensional design and program assessment and optimization are performed using digital chemical plants such as FDS (Factory Design Suite). Figure 5 shows the difference in design of two-dimensional and 3D layout.

Figure 5 Design from 2D layout to 3D layout design

After the modeling tools of the three-dimensional factory, such as FDS, Revit and other software, the modeling of the three-dimensional plant is completed, and the simulation can effectively avoid the spatial position of the plant structure, utility, and the like, as shown in Figure 6. Figure 6 FDS three-dimensional factory design

In addition, through the simulation of the construction process, the construction risks can be avoided in advance, and the construction efficiency can be improved. Through the flow analysis of the pipe, the utility of the factory can simulate the economics of the assessment plan. The modeling of the three-dimensional factory has laid a certain basis for the work of subsequent human engineering simulation. In order to further improve the efficiency of the assessment, the Virtual Reality, VR is introduced, and the assessment personnel can be used to visit and evaluate the factory that has not yet completed.

2.3 Production line simulation verification

After the production line plan, you need to assess production capacity, logistics, human machine, etc. to ensure planning plan. In recent years, with the maturity of virtual simulation technology, virtual simulation provides a strong means for pre-schedule assessment.

2.3.1 Production capacity and logistics simulation

The capacity simulation refers to the model using Witness or Plant Simulation and other software to model the actual production line, combined with all station beat (CT), mean time to recover, MTTR), average failure time (Mean Cycle Parameter settings such as BetWeen Failures, MCBFs, simulate the actual production status, verify the production line capacity, to achieve the minimum equipment investment to meet the planning capacity requirements, the specific interface is shown in Figure 7.

Figure 7 Simulation interface

Simultaneous simultaneously using the tool to complete the simulation of logistics operations simultaneously. In the logistics simulation environment, the input logistics scheduling strategy, the number of AGCs, the number of stereo library bits, and the operating parameters to simulate the planned AGC transport path to operate logic to ensure the overall design meets the demand. Based on the operation of the logistics system to meet the production needs, the leaving of the Optimization AGC number, path, and three-dimensional library bits are achieved by sensitivity analysis.

2.3.2 Dynamic assembly simulation

Dynamic assembly simulation refers to software in Jack or PD / PS, based on the actual operation content and the environment to model the modeling of the scene to analyze the operation interference, human machine, beats, etc. during assembly process. Figure 8 is a dynamic assembly simulation of a certain transmission assembly process. Dynamic assembly simulation has the following role:

1) Timely discovering the problem of product interference, reducing project risks.

2) Eliminate potential human machine problems in the early stage of factory planning, reducing rework costs.

3) Waste by optimizing simulation, reducing turning, walking.

4) Optimize the layout of the line and the tooling design, optimize the production efficiency.

2.3.3 Virtual Reality Simulation

In recent years, virtual reality technology gradually came out of the laboratory and began to use commercial and industrial use. The biggest advantage of virtual reality is that the evaluation person can be given an immersion in the previous design, which greatly improves the efficiency of design assessment. In the industrial field, virtual reality technology can be used in three-dimensional factory assessment, product process assessment, virtual assembly training, security, etc. Among them, the production line design link planned in the factory, the engineer roams in the virtual scene, evaluates the rationality of equipment, production lines, and plant.

Figure 8 Dynamic assembly simulation based on PD / PS

In terms of process assessment and safety design in the factory plan, the assembly process of the product can be simulated in the virtual reality environment, identify process issues and safety design issues. We speak virtual display simulation synchronization to training applications. The current product disassembly training has been applied in SAIC. As shown in Figure 9, it is possible to quickly become familiar with process operations quickly before the production line has not yet landed. In the equipment design link, the virtual reality technology can be utilized to simulate the repair of the device, evaluate the maintenanceability and equipment security.

Application of Virtual Reality Technology in Transmission Assembly Training

Figure 10 Application points of SAIC GM intelligent in factory planning

Figure 11 Intelligent Process Development in Process Planning

In terms of machining, the setup of traditional walking path needs to take a certain time with the experience of the CNC engineer and set debugging. With the intelligent arrival, we provide our new solution, Figure 12 is A typical example of a typical CNC tool path path. By automatically read the pore information of the product, combine the spindle motion parameters of the machine tool, use the intelligent algorithm - genetic algorithm, try geometric number of different walking paths, and evaluate based on rules such as walking time, quickly get the optimal Take the path to improve the efficiency of CNC programming and the production capacity of CNC.

Figure 12 CNC walking path diagram

3.2 Intelligent production line layout

Figure 13 is a layout diagram of a production line. The production line layout is a very cumbersome task, especially in the initial planning stage of the factory plan, often needs to optimize the layout. This application is currently in the research phase. The smart production line layout is to establish a device CAD model repository, including various forms of equipment layout. The device element is associated with the production capacity, energy consumption and other information, and the equipment CAD model is used as the layout element, with the needs of the field, production capacity, lean layout, and the layout constraints, and the geometric position of the equipment model as an argument, through intelligence The algorithm quickly attempts a variety of layout schemes, based on all kinds of lean indicators of the production line layout, thereby obtaining the optimal layout, which can greatly enhance the efficiency of the layout design. Figure 13 Layout diagram of the production line layout

3.3 Intelligent workers, human-machine analysis

SAIC's secondary development of the standard working hours, uses group thinking, and makes common continuous movements, changing our original selection parameters in the way to generate multiple records in a one-time generation, thus improving our preparation The efficiency of standard work hours. The concept of fusion machine learning will also be fused, and the input operation description is automatically pushed to our corresponding group records to further improve efficiency in the form of auxiliary form.

The future research direction is automatically generated standard work. The VR technology fusion action capture technology, simulates a station operation, background associated jack, or PD / PS such as a virtual scene, real-time automatic generation of standard work hours, as shown in Figure 14. Synchronization The human machine problem during the operation will be automatically recorded, which is easy to intervene, and eliminate the problem in time.

Figure 14 Intelligent worker, human-machine analysis principle

3.4 Dynamic Product Energy Simulation

Dynamic productivity simulation is an upgraded version of the production capacity, which emphasizes the interactivity between production data and capacity simulation software, and can be used in plant planning and operational links. The main modifications are as follows:

1 For the PMC (Production Monitoring & Controlling) data of the project after SOP (START OF Production), the data is abnormal, and the data visualization is visualized, reminding the field engineer to investigate the problem.

2 Analysis of future system capacity by historical operation, to optimize inventory.

3 In the output of the production capacity, the fusion machine learning model is used to eliminate the systemic modeling error of the production capacity simulation software, and further improve the accuracy of simulation. Currently, the first step in the application has been implemented, second, and three steps are being studied. Several relationships between several links are shown in Figure 15.

Figure 15 Dynamic production capacity simulation three steps

In addition to the exception of the above case, another important research subject to be considered is the intelligence of simulation software, and its characteristics mainly include:

(1) Project experience library. Working logs are constantly generated during the simulation project, which should be guided by subsequent work tasks. The previous work log and the simulation model step are retained in the form of a database, and the computer learning engineer will be quickly resolved at the help of the computer's assistance to the computers.

(2) Focus on simulation results and the actual results. At present, most software does not evaluate the modeling method of the engineer. For this problem, it is necessary to import the actual operation of the simulation object, import to our simulation log database, form a complete project closed loop, and implement objectively evaluation mode. Select the modeling parameters to optimize the simulation modeling standard.

The intelligence of factory planning is mainly reflected in the development of process intelligence, and the production line plan is intelligent. The trend of process intelligent development is integrated with unified platform, and is highly integrated with product design to lay the foundation for rapid derivative design. In the product design, the manufacturability and processing equipment capabilities of process development needs to be considered, thereby improving design efficiency and optimizing design. The trend of export planning intelligence is to use artificial intelligence technology to realize production line layout programs, and automatic production and evaluation of three-dimensional production lines and factory buildings.

(4) Factory planning prospect

Around the "fueling, efficiency", future plant planning will become more intelligent and accurate, thereby increasing planning quality and efficiency, enhancing the market competitiveness of enterprises. I believe that with the continuous deepening of digitization, intelligent technology, more creative technology will eventually achieve unmanned factory planning vision.

Author: SAIC MTAC Co., Ltd.