Signaling is where downstream and upstream processes are linked to the pacemaker by simple rules: First, every upstream process produces only when provided a replenishment requirement or an authorization signal from the next downstream process (pull signals such as kanban). Second, processing downstream from the pacemaker must occur in continuous flow (with the exception of the finished goods supermarket, if needed.) The result of this control system is that the pacemaker gets a schedule and every other internal and external supplier in the factory and supply chain is synchronized with the pacemaker.

Signaling can be accomplished in a variety of methods. Traditional methods such as cards or empty totes can of course be utilized, but fortunately leveraging technologies such as RFID or MES (manufacturing execution systems) add benefits of timeliness for real-time labor savings and accuracy. An integrated approach is vital to enterprise-level lean production management.

Signaling keeps production of all other processes in synchronization with the pacemaker based on actual execution throughout the shop floor. Ideally, signals provide work centers with a number of benefits:

Figure 2. Example of a signaling screen for coordinating SMT production.

Using these elements, lean production management can also be viewed as 'real-time execution-based scheduling' that keeps the factory activity in perfect synchronization with minimum effort.

For high-mix EA operations, 'the devil is in the details'
For factories that produce high variety or custom-engineered products, the calculations for pacemaker scheduling must also take additional factors into consideration, such as the mix of build-to-order (BTO) and build-to-replenishment (BTR) items on the same resource, frequent new product introductions (NPI) and end-of-life (EOL) planning.

In certain facilities NPI activities may be conducted on a dedicated line prior to ramping to production volume, and BTO types of production may also be handled in a similar way. Shared resources and lines are a reality in production that must be considered in providing an effective solution.

Instead of being 'inventory-driven,' where supplying processes are triggered to produce parts to replenish an inventory buffer, new signaling methods utilized in high-mix environments are 'capacity-driven.' With this new method, supplying processes are triggered to produce the next item needed when capacity becomes available in downstream processes.

Wherever possible, strategic planning for spare capacity on critical resources enables even greater system responsiveness. This may seem somewhat counter to 'running flat out' in the big picture-but the ability to satisfy the customer order is the critical point, not the100% utilization of a particular capital resource.

An important consideration of lean production management is controlling the level of material and WIP in the process in an effective way. Whether the signal technique is Kanban or another technique, the objective is the 'WIP cap'-that is to try to drive the process to the right level of material. Reduction of WIP also highlights hidden problems that, once visible, enable continuous improvement of the production system.

Figure 3. The classic 'WIP ship.'

The benefits of reducing the inventory are clear. How then does one implement a production management system in an environment where Kanban is not appropriate?
One method of production control is known as CONWIP [3] (CONstant Work In Process) where once a specific level of WIP is reached, no new jobs are authorized for the production system until work is completed and leaves the production system. Demand is accumulated on a backlog list if the system is 'full' and is released to the system when capacity is available. If the system is not full, work is released to the system, according to the demand signals received. Like kanban, WIP can be controlled by the number of 'cards,' but in the case of CONWIP, this card is not part-number specific.

Work on an item in the production system does not begin unless both conditions of available capacity signal and a demand for the specific item are met. If both are not present and available, production activity will not start.

Figure 4. CONWIP principle.

Once production activity starts on the item, the capacity signal or 'card' is not available until work on the item is complete. Tracking the execution on the shop floor with systems such as MES can provide additional benefits.

Another hybrid control system is known as POLCA[4] (paired overlapping loops of cards with authorizations). POLCA has some similar characteristics to CONWIP with 'cards' as capacity signals and the 'authorization' or 'backlog' list. Production activity will not start unless both items have been satisfied. Additionally, POLCA is very well suited to controlling the movement of products that must take variable routes through a production system with multiple shared work centers. Essentially the use of POLCA cards assures that each cell only works on jobs that are destined for downstream cells that will also be able to work on these jobs in the near future. If a downstream cell is congested, it will not release the POLCA card to the upstream partners. The upstream partner's capacity is ideally better utilized, if possible, to work on an item needed by one of the other available downstream work centers.

All of these methods discussed share the characteristic that if a work center goes down (for whatever reason), upstream processes will stop their activity when the capacity or replenishment signals have been 'consumed.' Further WIP accumulation will stop.

Value stream mapping
To understand the complexities of a modern electronics factory, especially with multiple production areas and product families, one must have a way to understand and visualize the material, product and information flow in an objective way. Fortunately such a tool exists-value stream mapping.

To be able to improve any system, no matter what it is, you must first have the ability to measure it. Additionally, once one 'has the big picture,' one can consider the application of appropriate technologies and operational practices in the future. Required improvement activities are highlighted

All too often in EA our attention may be fixed on the optimization of one particular process in isolation. This leads to driving practices in production such as 'Achieve 100% Utilization on SMT Placement' as a KPI measurement. This may make the individual department look very favorable-often achieved by running larger batches-but in the larger picture can result in unnecessary WIP or a loss of flexibility for the overall operation.

Figure 5. Isolated improvements.

Conducting the value stream mapping exercise
Preparation for a VSM exercise is very important. The first step is to find an internal project sponsor to act as champion and, in some cases, assist with some of the organizational issues-more commonly known as 'politics.' In selecting the team members for value stream mapping exercise, it is important that the make-up of the team includes various disciplines from the factory-production, engineering, planning, quality, material control, stores and management. The attempt is to involve a representative from each stakeholder department. Without team participation and commitment, 'silo thinking' may otherwise kick in, and there would be no way to reach a consensus on the results.

Figure 6. Value stream mapping steps.

To understand the operation of any EA facility, one must understand its purpose first, which very clearly is to assemble products according to customer demand. Understanding production demands is just the first step in any value stream mapping exercise. “Just what is it you do here anyway?”

Implementing lean production management
During the VSM exercise and in the measuring of the current state, process steps that require improvement, as well other improvement opportunities, are clearly recognized.

This identification process can lead to a “kaizen” or continuous improvement opportunity.

Kaizen opportunities identified can be specific to a process-'point kaizen'-or can relate to the entire production process-'system or flow kaizen.'

Improvement can begin right away on many of the identified opportunities with little or no capital investment. Realization plans can be worked out, and many of the improvements can be put in place in a relatively short time frame.

Moderate term improvements in some cases can be realized with software upgrades, with improved functionality and performance without major changes to the current business process. Redundancy of effort can be eliminated and, in some cases, where data is being collected manually, automatic web based data collection and reporting can be implemented, with labor-saving and reporting timeliness. Performance information serves as the basis for further continual improvements. “How can you improve what you can't measure?”

Fortunately for today's EA operation there are several technologies that are now available that can help in the realization of lean production management, as never before. While not all technologies are well suited to any particular operation, the VSM exercise does a good job in highlighting some appropriate candidates from these technologies for potential implementation:

• Rapid and error-free transfer of CAD & BOM data to product & recipe definition on the production floor. Systematic control to ensure the latest and most correct information.
• Empowering the production planner to create 'effort aware' production schedules and sequences to reduce effort on the floor while meeting deadlines.
• Use of group technologies to eliminate or minimize changeover, -including intelligent family setups clustering.
• Coordination across departments to synchronize material supply and preparation activities with the production line needs.
• Use of additional hardware and software to support external setup activities to enable rapid changeovers and smaller production batches.
• Use of online 'floating changeover' techniques to minimize material handling and maximize production line availability by reducing or eliminating changeover times.
• Enhanced visibility of resources and locations, including material and tooling, movements, completions and consumption.
• High production equipment availability through totally productive maintenance practice (TPM) and usage-based maintenance management software.
• Web-based real-time alerts and process KPI reporting to support timely actions and visual manufacturing, including Andon displays.
• Error proofing-machine tooling/material setup validation and automatic process interlocking.
• Comprehensive planning solutions to support BTR and BTO production strategies.
• Real-time signaling on the production floor to coordinate production job execution.
• Leveraging technologies such as MES, RFID and Internet-based technologies to gain visibility and integration into the supply chain and the manufacturing floor.

Conclusion
The deployment of TPS and lean manufacturing is not a destination but a journey. With an integrated approach, there are proven benefits to pursuing the path of continuous improvement. Fortunately today-within EA-there are tools and technologies to assist in the success of institutionalizing lean production management within the operation.

Dr. Tuan Nguyen is director software strategy at EA. Dr. Nguyen holds a Ph.D. in business administration with focus in operations research and operations management. Tuan brings 25 years experience in providing solutions for operational excellence in electronics manufacturing as well as past participation in the technical advisory board of the award-winning research project CIIMPLEX (Consortium for Integrated Intelligent Manufacturing Planning and Execution), funded by the DOC (Department Of Commerce) and NIST (National Institute for Standard and Technology).

Vern Harrison currently is software product manager at EA. Vern has a background in mechanical engineering and more than 20 years involvement in electronics manufacturing and SMT. Vern brings a strong operational background with varied roles, including process development, new product introduction, applications and project management. Vern's current focus is to bring software solutions to market, with emphasis on institutionalizing lean production management and operational excellence.