Streamlining Production and Device Programming Processes
December 31, 1969 |Estimated reading time: 9 minutes
By Jeff Williams and Scott Newell
Implementing computer-integrated manufacturing can tighten the manufacturing process.
Increasing time-to-market pressure continually presents high-volume electronics manufacturers with the challenge to streamline production. Recent advances in manufacturing equipment and processes means that many traditional production constraints no longer exist. Additionally, they present a practical solution to tighter process integration while providing manufacturers with the opportunity to leap forward in production methods.
Computer-integrated manufacturing (CIM) is key to taking full advantage of a streamlined process.
Market Trends Drive CIM
High demand for electronic products, from cell phones to DVD players, means that manufacturing is continually stretched for more capacity. At the same time, manufacturers want a lower cost per unit and faster turnaround.
Typically, these goals are achieved by adding production lines, infrastructure and facilities; increasing individual line capacity; improving yields; minimizing downtime; or automating the manufacturing process.
Cost reduction per unit can be achieved either by reducing infrastructure or increasing capacity without increasing infrastructure. A third option exists: increase process efficiency and remove overheads by further integrating production processes.
New advances, such as in-line device programming, are bringing processes typically performed off-line into the production line, streamlining the process, reducing lead times and production costs, and product customization.
The Manufacturing Process
While most manufacturing processes are well-defined and streamlined, many remain islands of integration. While standards exist that define the interfaces between manufacturing equipment, the challenge is to leverage them with standards used in other systems to achieve CIM at a conceptual level. Integration of the various processes must use the technology, i.e., local area networks (LAN) and software. This requires thorough planning and an understanding of the process, which in turn requires a definition of the integration level required, the type of information to be excha-nged and needed formats.
Figure 1 shows a block diagram of the high-level enterprise resource planning (ERP) systems that exist in the modern manufacturing environment.
Figure 1. An overview of the ERP system.
The ERP system feeds scheduling data into a CIM-based production line and extracts production analysis data. Scheduling data include the production schedule, material schedule and the data file required for programming. As the information is given to the production line, jobs are run and tasks are performed, generating production analysis data feedback. These data are fed back through the line and sent to the ERP. Production analysis data include inventory levels, completed and uncompleted jobs, orders filled, cost data, and yields. In addition to SMT equipment, an in-line programming system can be added to the CIM-based production line to integrate device programming.
Generic Equipment Model (GEM)/ Semiconductor Equipment and Materials International (SEMI) Equipment Communications Standard (SECS) is a standard that defines how data are passed between semiconductor manufacturing equipment and host controllers. SECS is comprised of two parts: SECS-I defines a communication interface suitable for the exchange of messages between semiconductor processing equipment and a host; SECS-II provides the definitions of messages and related data items exchanged between host and equipment. GEM is a standard for defining semiconductor equipment behavior as viewed through a communications link. This standard defines which SECS-II messages should be used in what situations, and what the resulting activity should be.
The SECS-I standard is used on older semiconductor manufacturing equipment. Newer equipment that requires high-speed communications or cannot be configured on a point-to-point topology network are adopting a new standard high-speed SECS message services (HSMS).
Bringing Processes In-line
Off-line processes in high-volume manufacturing environments present numerous issues, including cost and the inability to respond to last-minute changes quickly, such as when customizing products.
Whenever a manufacturer brings a process in-line, however, line balance must be considered. Bottlenecks always exist in the production process; they simply move from one part of the process to another, as each element is streamlined. For instance, if the test time for a particular printed circuit board (PCB) already is in the range of 40 to 60 seconds and the cycle time (assembly time for a single PCB) is 25 seconds, then two test fixtures are required to eliminate a test bottleneck:
40 seconds test time / 25 seconds cycle time = 1.6 rounded up to two systems.
By adding programming to test, a manufacturer may add 16 to 80 seconds to the test time. This results in a total test time of 56 seconds, requiring the addition of a test system or fixture:
56 seconds test and programming time / 25 seconds cycle time = 2.24 rounded up to three systems.
Table 1 shows the implications of programming and test time in relation to cycle time.
Production Process Improvement
By choosing the appropriate method to bring a process in-line, the line can be re-balanced, even streamlined. A key production goal is to increase the output from each line, either by decreasing cycle time or line downtime, or increasing production yields.
Table 2 details the impact of improving these parameters. The question is can this be achieved practically? A large part of the answer lies with CIM. Comparing the first scenario to the last, the implication is an increase in line output by more than 350,000 products a year, multiplied by the average number of production lines in a medium sized facility (6), delivers an incremental capacity of more than 2 million products a year. Assuming a value of $100 per unit (sales revenue), this is $210 million per year, without any changes other than streamlining the process.
So how can CIM contribute to this picture?
Cycle Time
Cycle time is determined by assembly equipment capability and balance. However, other process elements, such as test and programming, can be a bottleneck. The goal is to balance all processes across the line.
Assembly machine software has made tremendous advances in determining the most efficient algorithm for balancing assembly processes, i.e., which parts to pick-and-place, what order and on which machine. Unfortunately, assembly remains an island of efficiency. The ideal would be to apply similar performance algorithms to the combined processes, requiring tighter process integration using CIM. Depending on the job scheduled for the next shift, day or week, assembly, test, programming and other process elements could be balanced for optimal cycle time. This calculation could be performed manually, but this is a tedious and unnecessary approach, given today's technology.
Line Downtime
Line statistics can reduce downtime, or at least identify the weakest links in the process. CIM helps by capturing real-time information from different process parts and providing feedback on weak areas. Currently, much of this analysis is achieved by collecting statistics from various production stages and plugging the results into formulae or software located on the supervisor's PC.
The challenge is identifying the relevant information required from each production stage, and collecting and consolidating it. Systems exist to accomplish this, but actual implementation often is achieved through custom, home-grown systems, making implementation difficult to sustain as processes and needs change.
Ultimately, the central processing/control station should notify operators and supervisors in real-time of ongoing situations, alert suppliers and send them critical information (pre-specified), and use Web elements to solve the problem. This is not much different from the telecommunications systems that report their own faults back to the central control office and allow technicians and engineers to come online to troubleshoot the problem in real-time.
Customization Opportunities
Until recently, programmable devices were used in only the most complex and costly products, or when configurable memory was critical. But now, the consumer electronics market consumes the most flash-programmable memory of any industry.
Programmable flash memory devices allow designers and manufacturers to produce many product variants based on the same basic hardware by programming different codes. This enables just-in-time and on-the-fly manufacturing, opening the door to real-time product customization. This trend is just emerging, with many products having half-a-dozen variants in terms of models and capability.
The logistics of an off-line process to manufacture these products almost are inconceivable. In terms of inventory, cost and the risk of errors, this goal nearly is impossible without bringing the programming process in-line, to be just-in-time.
Imagine the day when the order is received and entered into the ERP system, triggering the materials order, job schedule and production process. The consequences on time-to-market and costs are substantial, especially when the manufacturer continually streamlines the process, improves cycle time, decreases downtime and increases yield. Products can be traced consistently and reviewed easily.
So, what will this next step in production process streamlining take?
Manufacturing Process Integration
Figure 2 offers a CIM process scenario. A great difficulty in integrating CIM and ERP systems is the number of different systems that exist. Many companies use systems customized to such an extent that an exhaustive amount of manipulation is required to successfully integrate the two. Other companies use a particular ERP system that is incompatible with prospective CIM systems. To force new manufacturing equipment types to interact with existing CIMs or ERPs, custom modules need to be developed. There are many roadblocks that can slow the integration process; these are some of the challenges to achieving true CIM.
Integration Example: Programming Process Control Software
Historically, programming has not been a mainstream production process. But critical programming information has led to the development of programming process control applications.
The inputs include the choice of specific programmable device, data file, programming parameters (which usually are device and data file specific) and production flags, such as the device quantity required, acceptable yield limits, etc.
The output is a correctly programmed and verified part, validated by a correct checksum and process statistics. System statistics also are provided that indicate system status and any system issues or failures throughout the process. These can be used to minimize system downtime, as well as allow the user to automate the programming process by defining and saving specific jobs that can run as required.
Figure 2. The CIM-based production line.
The next step is to integrate programming into the mainstream production process through CIM. The architecture of many software programs is modular and designed for added enhancements. The opportunity exists for production systems, including MRP systems, to select the correct job from a database, depending on the specific programmable device to be assembled on the next PCB. Likewise, selecting different jobs on-the-fly allows for customization. Without this capability, customization is not feasible.
Presently, the application is independent from the production process. Through LAN technology, standards, common operating systems and a thorough process understanding, programming (along with other critical production processes) can be incorporated to improve and streamline the production process, thus reducing lead times.
Future Directions
In addition to technology, a change also must occur with-in the business community. Value-added partnerships are cropping up throughout the industry, offering customers solutions to production needs vs. point solutions.
Standards emergence and adherence is the first step in supplier cooperation and partnerships. Suppliers will have to work closely with customers, perhaps compromising their solution to further integration, as well as working to understand the customer's big picture and production needs.
The technology and capability to achieve these goals exist today. With the right application, industry and supplier cooperation, and some innovation, someone will lead the market to the next level of factory automation through CIM.
JEFF WILLIAMS and SCOTT NEWELL may be contacted at Data I/O Corp., 10525 Willows Road N.E., P.O. Box 97046, Redmond, WA 98073-9746; (425) 881-6444; Fax: (425) 882-1043; Web site: www.dataio.com.