Program-at-Placement: Making Manufacturing Leaner

A lean enterprise uses time and the relentless pursuit of waste elimination as competitive leverage. It seeks to make value flow from raw material through consumption using the least amount of resources including time, people and materials. Successful companies adopt lean principles to create a culture of never-ending improvement at all organizational levels. This article focuses on the benefits of program-at-placement, a lean manufacturing process, and its impact on high-volume manufacturing.

The quest for putting the power of the personal computer in the palm of your hand continues to drive the need for higher density Flash memory. Internet-enabled devices such as smart phones are at the forefront of this growth. Flash density requirements for handheld wireless products continue to grow exponentially. Figure 1 shows how the increase in broadband access is forecast to greatly increase the amount of Flash memory used per cellular phone. As shown, the average number of megabits per cellular phone is forecast to increase at more than 50 percent per year.

The growing number of handset features continues to rise. As seen in Figure 2, the future of high-speed (that is, two megabits) data transmission capability belongs to advanced third-generation cellular services. The main benefits of high-speed data transmission will be the availability of video conferencing, web browsing and full motion video.

A major challenge facing today's high-volume electronic manufacturer is meeting production TAKT times (the pace of production required by the customer) while controlling costs. This is no trivial task when one considers the rate at which handset OEMs are adding new product features.

The Flash value stream begins at the semi house where raw stock (blank Flash), is manufactured, packaged and delivered to the customer. Products using Flash memory take on their unique character after the intellectual property has been programmed into the device. Intellectual property defines a product's capabilities via feature sets that make the product unique. Too often the intellectual property, often referred to as software code, is not finalized until just prior to, or shortly after, the production build is scheduled to begin. This jeopardizes the manufacturer's ability to deliver product on time. If the intellectual property is not ready, product may be built without the software code and stored in inventory. These inventories may eventually get reworked (re-programmed) and sold, or end up in the scrap pile as shipping delays evaporate the market opportunity. Both scenarios waste valuable time and resources, having a negative impact on corporate profits and market share.

Until recently, the two most widely used practices for programming Flash memory were either outsourcing to a programming center or in-system programming at test.

Outsourcing relies on the delivery of pre-programmed Flash memory devices from the programming center to the manufacturer. Holding pre-programmed inventory is risky as last-minute code changes make existing stock obsolete. Reworking pre-programmed inventory takes time (days) and money. This can be financially disastrous when trying to get a new product to market on time. As shown in Figure 3, the manufacturer can be left spinning in a vicious cycle, as outsourcing delays, due to code changes, slowly diminish the OEM's market share.

In-system programming was created as a result of combining the programming operation with in-line functional test to provide just-in-time programming. In-system programming was the preferred Flash programming process for high-volume production when board TAKT times averaged 20 to 30 seconds and 4 megabits of Flash was adequate.

Today, TAKT times are half what they used to be. Some products are pushing Flash density to 128 megabits and higher. How are today's high-volume production managers preparing to build these next-generation products? By adopting a new process called, "lean manufacturing."

"Program-at-placement," a lean manufacturing option for in-line Flash memory programming, is changing the way production managers think about balancing the production line, as seen in Figure 4.

Line balancing is no longer segmented between board assembly and functional test. Optimized productivity requires balancing the entire production build process encompassing both assembly and test.

To understand the benefits of program-at-placement one must first have a basic understanding of the Flash value stream. Figure 5 shows the dynamics of the value stream highlighted by the key business drivers along with their value stream characteristics. The two key OEM business drivers are market share and time-to-market. Today, many OEMs contract out their manufacturing to an electronic manufacturing service (EMS) provider. Their key business drivers are volume and cost.

Lean organizations use time as a competitive weapon, gaining the ability to do things for customers that their competitors cannot afford to do. They do so by attacking waste throughout the value stream, from understanding customer needs to raw "material" suppliers. Figure 6 shows the waste reduction of an EMS company that moved from outsourcing to program-at-placement. Gone is the off-line programming process and its non-value-added steps including reduced inventory, inventory tracking and complex logistics. Gone too are the off-line programming costs and rework fees. Time-to-market dramatically improves, as code changes are managed "just-in-time." These benefits apply to both in-system programming and program-at-placement.

In-system programming makes financial sense for programming lower density Flash. However, as Flash densities increase, so do the cost of programming at test. Production managers are again tasked to reduce expenses while increasing production output. Questions keeping many of today's high-volume production managers up at night include:

• Where will I put all the test equipment required for programming Flash memory?
• How can I convince my boss I'll need to buy more testers every six months?
• Where am I going to get the money?
• Which tests can I afford to sacrifice, as a last resort?
• From whom can I borrow more test operators?
• How am I going to manage board scrap?

This is where program-at-placement can be used to eliminate the test bottleneck by balancing the production line. In fact, program-at-placement can be used to compliment test. For example, 95 percent of the intellectual property can be programmed at assembly, still allowing the final configuration or five percent of the software code to get programmed at test. Substantial cost savings are realized with the reduction of test equipment and its floor-space requirements alone.

Program-at-placement uses an automated programming system that mounts directly onto the placement machine's feeder-bank, like a standard feeder. It requires no additional floor space, and should capacity demands exceed the capabilities of a single system, a second unit can be installed alongside the first.

Lean principles imply doing more with less and the elimination of non-value-added steps. Program-at-placement supports these lean manufacturing principles. Flash memory devices, provided from the semiconductor manufacturer, are pulled from receiving and delivered directly to the production line for program-at-placement.

Figure 7 shows an EMS company's simplified value stream using the lean principles of program-at-placement. The semi house delivers the Flash memory as needed. The OEM provides the intellectual property, and the programming equipment manufacturer provides the in-line automated programming equipment.

Program-at-placement in combination with advances in Flash silicon technologies, such as Intel's enhanced factory programming, are helping today's high-volume manufacturers increase profits with the adoption of lean manufacturing principles. Program-at-placement eliminates non-value-added waste, providing increased flexibility and improved "line balance."