Happy’s Essential Skills: CIM and Automation Planning, Part 1
There is a lot of talk and even information about automation, but I find that there is very little available on automation planning. This is one of my specialties. I started by getting a master's in EE in control theory. This went well with my B.S. in chemical engineering as I specialized in process control and IC manufacturing. This is one reason I took my first job at Hewlett-Packard; they wanted to automate the IC production, even back in 1970. I discovered then that there were a lot of companies offering equipment and software, but not a strategy on “How to Automate!” So working with HP Labs, we developed a ‘methodology’ that has worked for HP very well over the years. In my next column, I will focus on computer-aided-manufacturing and the connectivity issues with different protocols available.
However, the benefits will be derived only if certain cardinal principles are observed. This column outlines, briefly, the background of computer-integrated-manufacturing (CIM) and six of those automation principles: Superiority, Simplicity, Flexibility, Compatibility, Manufacturability, and Reliability.
Introduction
The consistent characteristic of successful application of automation in manufacturing has been the degree to which business and technical management understands and promotes the strategies, tactics, and philosophies used in modern manufacturing. By reviewing these philosophies of CIM, automation, management roles, mechanization, SPC, TQC, LEAN, MRP and Design for Manufacturing the successful implementation of automation in any company, small as well as large, will be enhanced.
Computer Integrated Manufacturing (CIM)
The strategies outlined here are considered CIM. HP was once in the business of selling CIM and called it “The Manufacturers Productivity Network” or MPN. The products were various software, computers, networking, interfaces and measurement systems.
At that time, HP had been in this business for a long time—longer than any other company. It started with government’s and others’ request for automated test and measurement systems. Because of the need to automate various measurement instruments/systems, HP created the first machine-to-machine protocol called “HP-Interface Bus” or HP-IB. This was later formalized into the IEEE-488 Communication Standard. HP needed an “Instrument Controller” and so purchased a unique 16-bit computer architecture from Union Carbide (unique in that the operating system was ‘real time and interrupt driven’). This became the HP2116 in 1964.
This computer was unique in that it had space for 16 interface cards for measurement instrument. What HP didn’t realize was that many companies were using the computer, not with instruments, but with multiple input terminals and printers, creating the first ‘Time-share-systems’. So HP created a smaller and lower cost version, called the HP2114. This led to the world’s first time-share system, the HP2000. Later, this would evolve to a multi-user business system called the HP3000 and used what they called the HP-Precision Architecture (HP-PA). Today, only six Enterprise Hardware architectures have survived (all American) the endless computer wars; HP-PA, IBM’s POWER architecture, Sun’s SPARK, AMD X86-64 and Intel’s XEON and Itanium.
The CIM Architecture was defined as early as 1980, when the Computer and Automated Systems Association of the Society of Manufacturing Engineers (CASA/SME)[1] published a presentation of computer integrated manufacturing in order to provide a common set of terms for its members. As published by CASA/SME, the CIM Wheel of 1980 definition was:
The ring surrounding the wheel represents various influencing factors (man and his degree of expertise as the human factor, productivity as the economic factor and computer technology as the technological factor) for the development of CIM.
The updated CIM wheel (1986) contained the following:
The wheel itself contains four functions, namely engineering design, manufacturing planning, production control and factory automation. If the individual functions are connected with each other and operate with a common database, an integrated system architecture is created which is represented by the hub of the wheel. This development has resulted in the realization that CIM, apart from factory automation and functions indirectly related to the operational performance such as design (product/process) and production planning and control, is also linked to common business administrative tasks such as manufacturing management, strategic planning, finance, marketing and human resource management. A further innovation was the addition of information resource management and communications between the different functions. Therefore, a common database alone is insufficient for achieving integration. The all-embracing nature of the CIM wheel reflects the idea promoted by CASA/SME that CIM has to be viewed as a concept embracing the company as a whole.
Figure 1: The CIM Wheel defined by CASA/SME[1].
The outer ring
The common business administrative tasks related to CIM are located on the outer ring of the wheel. They mainly form the connection of the company to the outside world in general. Data processing applications can be found in the most diverse areas. Most software systems applied in these areas were originally self-styled developments, which are increasingly being replaced with commercial standard software packages. Currently this software is installed primarily on mainframes. Overlaps of its functionality exist mainly with the software of the production planning and control.
The Inner Ring
On the inner ring of the wheel, the functions closely related to the operational performance of the company are situated. Data processing applications of the development and design area are computer aided design (CAD), simulations, analysis programs such as the finite element method (FEM) as well as drawing storage and management, for instance with the help of group technology (GT).
The types of data found in this area are diverse: drawings, technical specifications, and bills of material. In manufacturing companies, the data itself are often in considerable disorder. Frequently there are several types of part numbers, more than one group technology system, many kinds of bills of material, a number of different CAD systems each having its own sort of computer internal representation of geometric data, etc. The applied software rarely runs on the same hardware, resulting also in a large number of different hardware systems.
The second group of applications on the inner ring of the wheel is attributed to process planning and production planning and control. It comprises tasks such as routing generation, resource planning, material requirements planning, capacity planning, order distribution and supervision, but also the planning of quality assurance (quality process and resource planning). In the USA, software in the production planning and control area mostly runs on large client-servers, although the software itself is more often than not supplied by sundry software houses and not by the computer vendor. As in the common business administrative area, the software packages—which at least are integrated within themselves—have a modular structure and their single components can also be bought and applied. Therefore, a company rarely has purchased and installed all modules of such a package, which in turn frequently results in functional overlaps and data redundancy (one example: material requirements planning and purchasing systems).
The third group on the inner ring includes the automation of the manufacturing installations. Examples are robots, numerically controlled machines, flexible manufacturing systems and computer-aided measuring and testing methods. This area is characterized by the extreme heterogeneity of the systems involved, the diversity of which being much more pronounced than in the previously mentioned groups of functions. Another view is the CIM Hierarchy, as seen in Figure 2.
The hub
The statements made above have already shown that at present within these groups of applications there are serious impediments with regard to integration. There are few suppliers covering all three sectors. Therefore, little or nothing has been done by the suppliers with regard to interfaces, not to mention the integration of the various groups of applications. Information and communication management, represented by the hub of the wheel, which links everything, is intended to serve as the information management and communication control function between the single areas. It operates on a common, integrated database.
Figure 2: The seven arenas of a CIM strategy.
Key Challenges
Three major challenges of computer integrated manufacturing are defined in Wikipedia[2]:
- Integration of components from different suppliers: When different machines, such as CNC, conveyors and robots, are using different communication protocols (in the case of AGVs, even differing lengths of time for charging the batteries) may cause problems.
- Data integrity: The higher the degree of automation, the more critical is the integrity of the data used to control the machines (see Figure 3). While the CIM system saves on labor of operating the machines, it requires extra human labor in ensuring that there are proper safeguards for the data signals that are used to control the machines.
- Process control: Computers may be used to assist the human operators of the manufacturing facility, but there must always be a competent engineer on hand to handle circumstances which could not be foreseen by the designers of the control software.
Subsystems in computer-integrated manufacturing
A computer-integrated manufacturing system is not the same as a “lights-out-factory,” which would run completely independent of human intervention, although it is a big step in that direction. Part of the system involves flexible manufacturing, where the factory can be quickly modified to produce different products, or where the volume of products can be changed quickly with the aid of computers (as seen in Figure 4).
Figure 3: The CIM Hierarchy of related /critical systems according to Wikipedia[2].
Figure 4: The CIM hierarchy of related/critical systems.
CIM Architecture for Manufacturing
The CIM data architecture that will be covered in the next column on computer aided manufacturing (CAM) is seen in Figure 5. There are standards of computer interfacing and control available in industry. Some are designed specifically for automation and particularly for electronics manufacturing, fabrication and assembly.
Figure 5: The CIM software architecture for manufacturing.
Automation
Two truisms are becoming increasingly apparent in industry: technology is rapidly advancing, leading to more complex products, and more and more nations await this technology advance, resulting in competition requiring an increasing focus on product cost and quality. In a model developed for PCB fabrication, a new, important variable, complexity variable (C)[3], has been increasing steadily since the 1960s at the rate of an order of magnitude every 13 years.
Automation comes from the words AUTOMatic and operATION, and it is a strategic tool for controlling, managing, and directing a productive process by automatic means. It usually is complemented by product and technological innovations. As an engineering discipline, it can be accurately planned; it is mostly arithmetic, not propaganda. The chief ingredients in automation is adequate know-how and common sense.
The business and global factors behind the movement to automation are numerous, these are but four:
- Global competitive pressures
- Growing complexity of product and working situation
- Changing skill availability and job expectations
- Technology availability and its costs
What has not been clear to management is that automation is principally an approach to a company's future business strategy.
Management’s response to automation has usually been fragmented and reactive with numerous requests for new machinery, (now using more computers), new processes and procedures, and the resulting situation has been overlapping or excessive investment requests accompanied by additional staff, with the all too often result of inefficient or incompatible fabrication systems.
Working Definition of Automation
Automation in a working context means more than just automatic machinery. Machinery implies mechanization, automation also means the system information to direct and control the people, materials, and machines, or as coined by many, systemization[4]. Automation, then, is made up of two components, like a vector—the mechanization or material flow, and systemization, the information flow.
Mechanization Classes
Mechanization can be divided into six classes, which indicate the amount of sophistication of machines
and machine interactions with humans, rated as to percent of the work done by machines:
Systemization Levels
By the same token, systemization can be divided into six levels that indicate the amount and sophistication of information, blueprints, data, scheduling, and control that takes place:
Each level has an increasing percentage of machine/computer content to handling the information required to fabricate, schedule, test or move a product.
Automation Matrix
When both measures are applied to any activity in the process to tool or build a printed circuit, then an automation matrix is created about that work center. This matrix, as illustrated in Figure 6, allows for the current situation to be appraised (even if it's all manual) as well as future objectives and plans. It is quite common to make automation objectives a number of steps or phases, in this way allowing each step to be stabilized before the next one is taken. The automation matrix lends itself to this step approach.
Figure 6: Automation vector is defined as systemization and mechanization including ‘material handling’ between work centers and networking between work centers.
Management’s Real Work for Automation
Contrary to many popular beliefs, the real work in automation is just getting started. The time for management to start is now. In general, there will be five challenges that executives need to concentrate on if an automation program is to take root and flourish:
- Commitment to be the best
- Building the team
- Tearing down traditional barriers
- Knowledge of the tools and philosophies that create excellence
- Leadership to execute the "strategies"
Commitment
The first step is a commitment. More precisely, a shared vision is the most important step. The vision that must be shared is that of being the BEST, and creating a road map for achieving that major goal.
There will probably be changes, and some major changes are often best effected from the top—by the general manager—whether he is the head of a stand-alone company or of a major manufacturing division. Only he can make the long-term commitment to being the best. This vision is shared, because it will require others to be committed to manufacturing excellence and to engineering excellence.
Building the Team
Successful automation can only result from a professional team effort. How can we get everybody pulling in the same direction? Part of the answer is education, the sharing of information about technologies that are galloping out ahead of the decision-makers. And where the general manager thinks the team is deficient, either internally or externally, he has to hire people who can do the job. He also has to be sure to adequately train the whole team, including the production workers, who too often are left out.
The technical demands may require a boost, at least temporarily, in engineering manpower or consulting. In many cases, the lack of adequately trained numbers of engineers limits our rate of improvement and increase the risk of failure.
Tearing Down Traditional Barriers
To build the team that is required, the general manager has to step up to the tasks of tearing down the traditional walls that have grown up and isolated the various functions that participate in manufacturing and engineering. He must build a strong partnership of equals: marketing, R & D, purchasing, production, engineering, manufacturing, sales, distribution, and after sales service—all the related functions working together as a closely knit team to achieve the overriding goal of being the best. These walls are barriers to not only that shared vision, but to the understanding and consensus of what information is needed to develop the strategy for automation.
Automation is Strategic!
There are numerous dimensions to the strategies in automation. All are driven by top management. Do you know any of these strategies? If not, then here are six of the more fundamental ones:
- Getting started—developing the plan
- Awareness of the opportunities
- The quality paradigm
- Recognizing the myths
- Understanding the prerequisites
- Avoiding the pitfalls
Automation is Tactical!
Likewise, there are numerous elements to the tactics in automation. If you are not aware of some of these, then here also are six:
- Focusing the factory
- Technologies to consider
- Implementing a manufacturing management information system
- Justification—learning to ‘pay as you go’
- Planning the steps/having a methodology
- Understanding how to integrate
Next: CIM and Automation Planning, Part 2: Six Principles of Automation
References
- CIM definition by CASA/SME.
- Wikipedia, computer-integrated manufacturing.
- Holden, H.T., "Complexity Factor C,” IPC Technical Review, March/April 1986.
- Wu, Bevan P.F., "Manufacturing Strategy Towards Integrated Automation," Taiwan Productivity Center Conference, December 1983.
