Facilities Design the Facilities We Plan Today

Facilities Design

The facilities we plan today must help an organization achieve supply chain excellence (Tompkins, et al. 2003, pg 3). Product design activities begin with conceptualization, where ideas for new product are generated based on market information or from existing technology. Product selection restricts new products to those ideas which pass the tests of market potential, financial viability, and producibility. In many cases product selection analysis may be quite subjective in nature and based on somewhat limited information. Once a new product idea has been selected for implementation, a preliminary design is developed. During this stage the product is specified completely. As part of the process many tradeoffs between cost and product performance are considered. A preliminary set of drawings may also be completed. The preliminary design is then operationalized as a prototype. The prototype is tested in an attempt to verify market and technical performance.

Several iterations through the prototyping process may be necessary. Engineering changes initiated as a result of prototype testing are then incorporated as part of the final design. During the final design phase, drawings and final specifications are developed. Design reviews are typically conducted after the preliminary design and final design stages during the design process. These formal reviews are aimed at determining whether a proposed design will perform successfully during use, can be produced at low cost, and is suitable for prompt, low-cost field maintenance. Usually the reviews are conducted by a team consisting mainly of specialists who are not directly associated with the development of the design. These specialists are generally in great demand and short supply in the organization (Shelton, 2008. Pg 42-43).

CASE STUDY I -- The Northrop Corporation:

Facilities Design Analysis

At a company like Northrop, the choice of technology deals with the set of processes, tools, methods, procedures, and equipment that will be utilized within the process type that has been selected. Once these issues have been resolved, the focus turns to two micro level decisions -- process flow analysis and facility layout. In process flow analysis we deal with both material and information flows from input through transformation to the final product. As a result of this analysis improved methods or procedures may be discovered. Based on the type of process selected and flow patterns developed as a result of process flow analysis, a facility layout is arrived at. These micro-level decisions affect decisions in other parts of operations including scheduling, job design, inventory levels, and quality control procedures.

As is the case for product design, consulting, design, and planning expert systems can be useful in supporting process design decisions. Consulting systems can be useful in advising designers on the links among corporate strategy, manufacturing strategy, process selection, and choice of technology. Design systems may be useful in determining the combination of parameters that will minimize the chance of manufacturing imperfection. They may also help in removing special causes and setting the process capability. Planning systems, again, may help in rationalizing the complete design process. Many of these process design activities fall under the responsibility of industrial engineering. Council on Competitiveness (2006), provides an interesting review of potential uses of expert systems in industrial engineering. Being a massive manufacturing company, Manufacturing Process Planner is a system developed for internal use by the Northrop Corporation that aids in the planning process for the manufacture of the approximately 20,000 parts that go into a fighter plane.

In control of quality during manufacturing consulting, control, debugging, diagnosis, interpretation, monitoring, and repair systems may prove useful. Such systems already have been successfully used in industry. As product difficulty grows there is likely to be an advanced incidence of field harms. Franz, et al. (2002 pg 2159) believe that 20 to 30% of problems concerning fitness for use are attributable to field factors such as inadequate operating or maintenance procedures, human error during maintenance, inaccessibility to repair, and defective spare parts. There exists here a great opportunity to reduce these problems through the use of expert systems in training and consulting capacities. Training systems can be directed both at service technicians and the user. Consulting systems can be used for the diagnosis and troubleshooting of operating problems. Many applications of this nature are already in place, including systems to diagnose steam turbine generator problems marketed by Westinghouse; a Toyota system to troubleshoot automobile engine problems; and COMPASS, a GTE developed system for assisting switch maintenance personnel by analyzing operating data and recommending appropriate maintenance actions.

The effect of shorter lead and cycle times can be evaluated within the company's broader supply chain perspective. This is the final step since the inventory, materials, and manufacturing infrastructures must be substantially in place to support the company's, suppliers', and customers' elements of the overall cycle time. However, in actual practice, policy, variability reduction, and cycle time reduction may be implemented in any order or mix that makes business sense.

At Northrop, whereas models provide snapshots over time, simulations provide a dynamic operational view of the company's material handling and storage processes. Plant- wide simulations begin with a CAD layout of the facility. Including elevations of storage systems, conveyors, and equipment enables a 3-D effect to be produced. Actual vehicle accelerations, operating speeds, decelerations, blocking effect of traffic in aisles, work station delivery and pickup distributions, and operator or automated load/unload times can all be simulated. Changing the operating variables, altering flow paths, volumes, and rates within the facility, and changing the number or assignments of operators can be evaluated in order to fine-tune lay- outs, equipment, and personnel levels (Anderson et al. 2009 p 169-170).

Northtrop uses simulations are developed in situations where the number and type of equipment may be in question, especially with regard to future upgrades and modifications. Production lines, buffer staging and storage systems, and material handling systems operating in constrained environments and subject to variation due to operator or other factors are likely candidates for simulation. A common application is to determine the number of automated guided vehicles required to de- liver unit loads to work stations located throughout the facility.

Companies such as Northtrop commit millions of dollars in facility layout modifications, new construction, and capital equipment investment on the basis of analyses, models, and simulations. Their product line being billion dollar products justify the high costs of facilities design. Management commits this level of investment only when convinced that the conclusions and recommendations are based on a valid representation of current and/or future conditions. While the decision-makers may not take the time to completely understand the actual mathematics and software logic, they do need to be confident that a structured methodology has been followed. They must trust the modeler. Confidence in the modeler is complicated by the fact that no two people will ever perform an analysis or develop a model in exactly the same way.

This can lead to different results based on the same requirements. Following a structured methodology is not so much to guarantee that the same results will be achieved with the same requirements, but that a valid result will be developed. Anyone who ever developed a bid spec and solicited proposals for an integrated system will attest to the fact that there is no such thing as the single right solution. Even with differences, there should be a range of outputs where modelers agree (Manufacturing Systems, 1993, pg. 104).

They should certainly agree on any aspect of the model which can be verified against current operations. If there are widely divergent conclusions and recommendations based on the same requirements, something is probably wrong. Having and using a structured methodology enables results to be understood, repeated, and evaluated (Shimokawa et al., 2007). The importance of graphing relationships pictorially via figures, tables, charts, and graphs cannot be overemphasized. Decision-makers trust a modeler's conclusions and recommendations more when the mathematics are translated into a visual format. This is one of the reasons why graphics-based simulations are such powerful tools. Another commonly used technique at Northrop, when developing System Requirements Documents (SRDs) and bid specs is to represent flow rates via block diagrams as shown below.

Flow in Figure 1.1.

This is a rather sterile treatment of flows that is utilized at the Northtrop Corporation and does a disservice to those readers who are not intimately familiar with the facility and analysis. To make matters worse, separate pages are often used to segregate full from empty moves and carton, tote, pallet, and rack moves. Over- laying the same flows on a facility layout as shown in the Facility Flow version makes a lot more sense. In fact, providing a feel for move distances and from/to points highlights the long moves and any reverse flows.

CASE STUDY II - Toyota Motors:

Facilities Design Analysis

Production is the heart and soul of Toyota and its precision guided, Facilities Planning initiatives. All Toyota lines are exactly alike, and it is impossible to mistake any one of them for a non-Toyota assembly line. Of course, cars sitting on the line are different and so is the appearance of the team members working on the line. The Toyota plants also vary in conspicuous features such as lighting and layout, giving different surface impressions. But something called a TPS (Toyota Production System) is also the unmistakable identity of all Toyota plants (Shimizu, 2004; pg. 230).

Inside the assembly plant TPS is embodied in both the physical realities and team members. In terms of physical realities, this relates to the way cars are carried, the way information is conveyed, the way parts are stored and fed, and the way machines are designed and function. Moreover, the latest additions to this landscape, such as the synchronized dolly and the raku-raku seat, are everywhere. In this regard, all the Toyota plants are extremely well standardized and there is a mechanism that assures they stay that way. In terms of team members, this relates to the way they approach their job, the way they work together, and the way they think. It is quite astonishing to hear all the team members speak in the same 'language' in different countries.

TPS encompasses an implicit contract between the management and the team members.

Those who have signed on as team members, with one minor exception, regard this contract as fair, and enjoy a sense of family or teamwork in the workplace to which they are inducted by buying into the common contract. The unique physical identity of the Toyota plant is a shadow of the implicit contract which expresses the ways in which management and team members fulfill their responsibilities to each other. The andon cord is a common fixture in every Toyota plant because it is part of the support that the management provides in order to facilitate communication, experiment, and problem-solving among the team members. Physically, it is a cord running above the work areas of the production line, or similar devices that are easy to operate while working (Shimokawa et al., 2007; pg. 160). A cord that runs over a few processes is connected to a toggle switch, which, when turned on, starts the matching electronic chime sound throughout the line, triggers a flashing lamp attached to the switch itself, lights up the matching section of the andon board as explained below, and stops the line when the cars on the production line reach the end of the processes they are passing (Shimizu, 2004; pg 67). A pull of the cord turns on the toggle switch and another pull turns it off. The first pull is usually initiated by a team member on the line, and the second pull is the duty of the rovering team leader unless the cord is pulled back by the team member who solved the problem himself. If team members find a problem, they can pull the cord above their head to call a team leader and let her know. If their process design is too ambitious, they can pull the cord to stop the line and thereby take more time per vehicle. If they want to rebalance the line, they can look at the number of andon pulls in each process as recorded by the conveyor control system and find out who is stretched and who is not. All team members agree that they cannot imagine a car plant without the andon cord. At TMM1, the assembly shop alone registers a few thousand andon pulls per shift, and the line actually comes to a halt a few hundred times.

The andon board, by indicating the status of each process, lets the team members know how their line is doing. This is a big board hanging over an aisle between the line segments or the line itself, and indicates the numbers corresponding to all the toggle switches in a group. Whenever the andon cord is pulled, a matching number lights up in yellow, and, when the line comes to a halt, the light turns red. The rovering team leaders (there are normally a few in a group of about eighteen team members) always pay attention to the andon board and spring into action whenever a light comes on, so that their line keeps running. The line, however, stops every now and then because the team leaders cannot solve all the problems that caused the andon pulls in time, or else they cannot attend to all the andon pulls occurring at the same time. When the line does stop, team members infer from the andon board what is causing the line stop and make an intelligent decision as to how best they should use the downtime: rest briefly, fill their parts bins, housekeep their work area, or go to help. By letting the team members know what is happening on the line where they cannot see other people deployed along a long line, the andon board creates the atmosphere of a real team (Shimizu, 2004; pg. 7).