Case Studies

The competitive aerospace industry requires companies to make changes to their manufacturing systems quicker and more efficient. Within the Boeing Company, the principles of lean manufacture are currently being implemented in all shops and processes. Flow time is being reduced, waste is being eliminated, and new processes are being put into place. The attitude throughout these shops is "Do it faster, better and more efficient".

The Everett Wire Shop is one of the areas in which lean manufacturing is being applied. Engineers have been assigned to look into designing work centers that will reduce cost, labor, inventory and flow time.

ProModel simulation has been used to assist this process. ProModel allows the engineers the opportunity to study, maintain and optimize the manufacturing cells. The model allowed the engineers to change operating parameters such as: part scheduling sequence, assemblies priorities, workable shifts, operator availability by shift, operator efficiency by work assignment, part processing times, kanban sizes, and operator priorities schemes.

ProModel has provided a tool that has

• Reduced the time to design the cell by 50 - 60%
• Helped shop personnel to maintain production as build rates changed
• Optimized cell performance.

The positive results from the first project in the Wire Shop have caused management to encourage and accelerate the design and implementation of additional cells. ProModel is currently being used to assist with the design of cells in wing sealing, skin milling, and shim cutting and has proven equally beneficial in these projects.

Production Modelling a leading UK based provider of simulation, planning and scheduling software and consultancy, would like to congratulate Burbidge & Son Ltd. on its exceptional application of the ProModel Simulation Software, for which it has also been recognised, as a highly commended case study, by MCS magazine.

With all businesses continually striving to meet new market demands and improve performance, change has become a way of life. However, trying new ideas out in an operational environment - be it the shop floor, warehouse or office - can take time, be disruptive, and be very costly, especially if they do not work first time. Similarly, few companies can afford for new equipment investments to fail to deliver the anticipated benefits. The solution, which is now being adopted by companies of all sizes, and from industries as diverse as general manufacturing to pharmaceutical and food processing, is to first experiment using ProModel.

Burbidge, an established manufacturer of high-quality kitchens and SME, has found that using its ProModel system to fully assess proposed production processes changes, and investigate investment decisions, delivers very real benefits; eliminating many of the potential risks and significantly improving decision making. And, despite ProModel being one of the most advanced systems in the field of dynamic simulation modelling, the software has proven extremely cost effective and easy to use.

As Graham Heaven, Burbidge's financial director, observes, "Too often companies still make process change or investment judgements based on gut feel, or extremely simplified and inaccurate assessments, and then wonder why these changes fail to meet expectations. Yet, as we have found, the tools for enabling important operational decisions to be based on realistic data are now readily available, easy to use, and require only a relatively small investment - especially compared to the cost of getting it wrong."

It doesn't matter what size of company they are, simulation should be an integral part of their decision making process, whether it's operational – e.g. adding new machines, changing layouts, extending the factory or warehouse, introducing new products, introducing new production methods, re-engineering a process, etc., etc. or strategic – e.g. can a capital investment be justified, will it produce desires performance improvement? How can we streamline our supply chain?

The cost of simulation is minimal compared to the benefits, and rapid return on investment, and increasingly, companies cannot afford to run the risk of making bad decisions. It's common sense really – you have to ask - why wouldn't you use simulation before making operational and strategic decisions?

The company in this case study, part of the GEC group of companies, proposed to reorganise the layout of a manufacturing facility at an estimated cost of £250,000. The aim was to improve efficiency and cope with anticipated increases in sales for two major product types.

The primary objective for the ProModel simulation model was to prove that the reorganisation would deliver the anticipated benefits.

For validation and benchmark purposes, the first task was to build a model of the current manufacturing system prior to simulating the proposed changes. This proved to be an extremely valuable exercise in its own right. ProModel helped us to communicate the ways in which both the current layout and the new layout were intended to work. It became clear that different members of the design team had different ideas although they were using the same words to describe the process. ProModel also promoted a greater understanding of how effective the current batch control rules were.

The simulation demonstrated that no significant improvement would result from the reorganisation of the layout alone. Additional improvements to the manufacturing system were required to realise the benefits of a cellular layout. ProModel clearly helped us to avoid a costly mistake.

ProModel's visualisation and ease of modification enabled the cell team to make and incorporate new improvements into the cell over and above the envisaged layout changes. A key component was the adoption of a Kanban system for production control.

The changes proved able to cope with the forecast sales increase and gave rise to significant improvements in product throughput. A 30% reduction in lead time was the most significant business advantage for the company.

In looking to help justify a major investment, Corus Construction and Industrial has found that the ProModel simulation software, from Production Modelling, has proved to be a vital new decision tool.

For all manufacturers, making the right investment decision has never been an easy task. But, with the prevailing business climate, selecting the best options in terms of new plant and equipment has become increasingly daunting. No longer can a company afford to assume that new technology will definitely deliver improvements, and any mistakes can quickly prove far costlier than the actual investment.

Therefore, it has become critical that prior to any investment decision being taken, both the financial returns and quantifiable performance improvements are determined as accurately as possible, and the case for investment accepted at every level. One way to help achieve this is to experiment and evaluate new equipment on a computer model before making the investment for real. This type of simulation not only provides the ability to measure the effects of new equipment on key performance criteria, and so help quantify financial and performance improvements, and can also be used to effectively compare alternative investment options. As such, this approach can take much of the risk out of any investment.

Yet, most companies still fail to take advantage of the benefits that simulation offers in terms of investment justification. The wrong assumption being that its application requires highly sophisticated modelling and so would prove far too expensive an approach, especially in terms the consultancy and IT skills needed. However, as Corus's Teesside Beam Mill has discovered through its use of the ProModel simulation software, even with just a few days training and the development of relatively simple models, it has been possible to very cost effectively analyse a major new equipment proposal. As a result, the operation has been able to determine that the investment will deliver the performance required, while also finding a solution that will cost less to implement.

New Evaluation Tool

The Teesside Beam Mill is part of Corus Construction & Industrial, one of the main business units within the Corus steel group, and is a major supplier of structural steel to the world's construction industry. Past investments enable the mill to compliment the rolling of standard sections with the production of innovative asymmetric beams, extra heavy jumbo sections and specialist sections designed for use in areas prone to seismic activity. Its products can be found in some of the most celebrated structures around the world; from providing the vertical columns in the UK's latest tower office building, at One Churchill Place, Canary Wharf, to being used in the rebuilding project at New York's Ground Zero.

Like all businesses competing within the extremely fierce global steel market, the Teesside Beam Mill is continually looking for ways to gain competitive advantage. While high quality, specialist products and co-ordinated customer support are one way, it is equally critical that the highly sophisticated operation continually improves its own productivity and delivery reliability. It was to help with investigating internal process and practice improvements that the site initially decided to invest in the ProModel Software, from Coventry based Production Modelling.

However, as Chris Hamlett, Manufacturing Manager - Development at the Teeside Beam Mill reports, "Having brought in the software and undertaken basic training, it was realised that the tool could also be easily used in a far more immediate role, to help evaluate a critical capital expenditure decision."

He explains, "The funding for major investment in the re-organisation of the operation's final product marshalling and dispatch activities, to further improve efficiency, were undergoing final approval. But, with theses activities being so critical to mill efficiency and with no chance of going back once the changes were introduced, there were significant risks involved. What we needed was a relatively easy way to more effectively analyse all the issues involved, and so ensure that the new process and equipment would not only meet all the demands, but actually deliver the improvements required. This, we recognised, was what ProModel could provide."

Mike Straiton, technical director at Production Modelling adds, "There is a long list of horror stories highlighting the significant risks involved in major production investments, and all point to the same failing - the lack of comprehensive justification analysis. With this in mind, we are coming across a rapidly increasing number of companies who like Corus are not just using ProModel simulation to improve existing facilities, but to help plan, check and justify - in detail - major new investments."

New Process Analysis

At present the mill employs two stacker cranes and a transfer bogie system to move finished products into the despatch bay. The product is then transferred to road trailers using three 15 tonne payload overhead cranes. The main issues with the current process are :-

• The 15 tonne crane payload is inadequate for modern transport needs.
• Despatch Bay inefficiencies as a result of crane interference.
• Current crane operation is time consuming and extremely manpower intensive.
• The current cranes are 1950's vintage.

When Hamlett came across ProModel, his team had already determined a potentially far more efficient marshalling and dispatch process. This involved the development of a pallet-based stacking system, modified trailer fleet, enhanced Despatch Bay coordination and crucially, a single modern large capacity crane based upon proven pallet / container technology. But, there were two fundamental questions that needed to be answered, namely :-

• Would the single crane cope with current production rates ?
• Would the downstream process cycle-time be maintained with the new palletised system ?

Therefore, the initial step undertaken by Hamlett, who was one of three trained in the use of the system, was to create a dynamic simulation model of the planned process re-configuration. "Despite our limited training on this advanced system, its ease of use meant that we were able to relatively quickly produce a simple, but more than adequate model for our purpose, and without the need for extensive outside help and consultancy," notes Hamlett. Although, as he adds, "The ongoing relationship with Production Modelling was vital as they provided feedback and important suggestions that enabled us to ensure that the model was accurate and fit for our purpose."

This simulation, driven with existing production data, proved a major step forward for the investment justification, by confirming that the new configuration and process sequence would work, and deliver the expected improvements. Moreover, the team were also able to use the model to establish that developing a new and potentially expensive transfer bogie system was not necessary, and that the required transfer process could be achieved by modifying and extending the existing bogie system. "Through this first model we not only established complete confidence in the basic premise of the planned investment, but we were also able to take a significant cost out of the project," claims Hamlett.

Then, after further development of the model, the team was able to undertake a far broader analysis of the new process, and in particular address the second key issue, which was the new crane. With the inherent reduction in flexibility of a single crane, the model was used to simulate a wide range of scenarios that assessed the key production parameters of movement cycle times, and product arrival rates. Again, the simulation clearly clarified the new equipment's ability to effectively cope with all current and projected throughput.

In addition, the development team now also have detailed output from the simulations that will be provided to the suppliers of the new crane installation. This is seen as another 'fail-safe' way of ensuring that the equipment installed will be exactly to the right specification.

Hamlett concludes, "ProModel made a major contribution to the development of the fully detailed re-organisation plan and its financial justification; for which we have subsequent gained the go-ahead. While the models developed may not have been the most sophisticated, they helped us build a far better and realistic understanding of the planned process, and provided clear and detailed answers to our key concerns. As a result, the project risks have been cut significantly, the cost of the re-organisation has been reduced, and all this has been achieved at a minimal cost."

Summary

Printronix was losing business because of our inability to meet the customers shipping requirements. Our solution to faster shipping was, with the aid of an analysis using a ProModel simulation model, to build an inventory of high level generic "printer plains". This inventory of generic "printer plains" enabled us to configure and ship printers within three days of the order.

Case Study

Printronix standard manufacturing process was to build a printer from the ground up at the time of order. Average lead time on a printer from time of order to time of shipment was 18 working days. The customers expectations were to have their order shipped within a week. In an effort to meet the customers shipping expectations, we built and inventoried finished printers based on a master scheduling forecast. We found that the mix and quantities were not always adequate to meet customer demand. In order for us to achieve 100% customer satisfaction we would have had to triple the existing finish printer inventory.

We decided to build a printer to it's highest generic level and hold it in a kanban. Upon receipt of the order, the printer plain could be quickly configured and shipped. This was accomplished by splitting the printer manufacturing process into two stages: printer plain assembly and final printer configuration. We used computer simulation to model this process and determine how many of the five printer plain assemblies would be necessary to support the 110 top level printers.

The results of the simulation model lead us to a printer plain kanban size that was 60% less costly than inventorying finished printers. We went ahead with the changes and the results were almost exactly what the model had predicted.

Lead time was reduced from 18 to 3 days, finished goods inventory was eliminated and a floor controlled just-in-time delivery process was implemented.

Siemens, DuPont, IBM, and Case Corporation were amongst those companies presenting supply chain case studies at the ProModel users conference, August 1997.

These are just some of the ProModel user companies whose customers

• expect greater responsiveness in fulfilling their orders and /or
• require highly customized products made or configured to order.

To stay competitive, the performance of the supply chain that includes both demand satisfied from inventories and demand for specially configured products becomes more critical. Companies are therefore using ProModel to evaluate strategies for reducing response time to customers and for shrinking the finished goods inventory along the supply chain.

Case studies show that ProModel can consider the entire supply chain beginning with the supplier, extending through the production and storage areas, and ending with the customers at several distribution locations. Based on those elements, ProModel is able to demonstrate the effects caused by changes on the demand patterns, the logistics control system, the level of safety stocks, the reordering algorithms, or even the structure of the supply chain itself.

For example, a supply chain may function on a pull or a push principle, work with low or high safety stocks, have different levels of distribution. A company may make some goods, and buy others. A company may be considering a relocation of a manufacturing site.

Among other capabilities, ProModel captures the random variations in sales, transportation lead times, supplier issues for specially configured orders, and inventory levels. By modelling the plant operations in sufficient detail in terms of major operations, product mix, and line scheduling, ProModel also provides a tool to balance supply and demand at an aggregate level.

Reported benefits of supply chain simulation using ProModel are significant and include:

• several $million dollars removed from an authorised project
• reduced operating costs due to fewer truck rentals
• customers have fewer incidents of running low on stocks
• cut inventory by 30% without reducing customer service levels
• capital cost avoidance of 3.5 million dollars.

The user group case studies also demonstrated ProModel's ability to model the transport fleet. In most real-world situations, this problem is highly dynamic (affected by planned shut-downs of producing and consuming units, seasonality of demand, changing products and customers, etc.) and stochastic (affected by random outages, variable rail transit and dwell times, variable rail car maintenance times, etc.). Hence, the fleets are usually sized conservatively (too large), to compensate for the many uncertainties. ProModel is an ideal tool for this problem because of its ability to handle the complexity, dynamics, and randomness. Some fleets have been dramatically down-sized as a result, with no loss of service.

Process analysis and system simulation tools are used by many banks to help operations managers improve customer service quality and resource productivity. These tools take into account the effect of variability, uncertainty and interdependency between customers and banking resources in various transaction processes. Union Bank of California used ServiceModel to improve their branch operations.

The Problem

One of the applications developed was the branch office simulation. It was developed to analyze the critical service elements of the branch including walk-in tellers, drive-up tellers, loans/new accounts service representatives, and ATM services.

The Solution

Union Bank of California developed the model of their branch operations. Animation was developed to visually show how long customer lines got under different operating scenarios. The on-screen statistics showed how current and proposed employee schedules handled variable transaction volumes.

The model demonstrated the effect of cross-utilizing the entire branch staff for customer and paperwork processing. The graphical and tabular output reports showed the effects of opening and closing teller windows based on the number of customers waiting.

The Results

The model was created as a template to be used for the entire branch network throughout California. Specific elements such as ATMs, drive-up tellers and employee schedules can be tailored to meet the needs of each branch office.

Using ServiceModel software, Union Bank of California demonstrated the benefits of specific scheduling practices that had been proposed and were under consideration. By implementing these scheduling practices, the bank will optimize branch operations.

Simulation Modelling, undertaken by Coventry-based Production Modelling http://www.simulation.co.uk, has played a key role in both the successful delivery, and the on-going management, of an expanded warehouse and distribution facility for Walkers Snacks Foods.

A rapidly growing snack food market, and the ability to capture an increasing share of this market, in part thanks to some high profile marketing, has seen Walkers Snack Foods Ltd experience major business growth over the past few years. But, this success has also placed considerable new demands on its UK operations. To manage these, the company - which is now part of the global giant PepsiCo Inc - has undertaken a major programme of upgrading and expanding of production and distribution facilities. One of recently completed key projects in this programme, a vital element in the company's remodelling of its distribution capabilities, has delivered a massive warehouse expansion at Walkers' Leicester based factory and warehousing complex.

However, adding storage was only one of the requirements for this development. It also had to deliver operational process improvements in terms of picking efficiency and throughput, and so enable the output capacity of the whole Leicester complex to keep pace with requirements. Therefore, to help ensure that the final design solution was capable of meeting these demands, Walkers, and its project partner, the warehouse solution provider Swisslog, turned to simulation modelling, and the UK simulation specialists, Production Modelling. As it has turned out, not only did the modelling help prove the design concept, it provided detailed process analysis that led to direct project savings. Additionally, a detailed simulation model was also developed that is now being used by Walkers as an ongoing operational management and training tool.

Project Background

By the end of 2002, Walkers had identified that to keep pace with continued UK sales growth, the warehousing and distribution capacity at its Leicester complex needed to be radically increased.

At the time, the facility included two factories with seven production lines, two of which were recent additions, a single warehouse, fed automatically by the two factories, and the Southern Region Distribution Centre (SRDC). Sited side by side, the warehouse and the SRDC, had previously been integrated through the introduction of a monorail system. This runs on a circular track round the SRDC, transporting full pallets from the high bay warehouse to either the SDRC's marshalling lanes, for full pallet pick, or to the picking frames, and also transport case picks from the picking area to the marshalling lanes. But, with the increased production from the new lines, the storage capability of this facility, at 9800 pallets, was far from sufficient. In fact, Walkers was having to store products off-site and bring them back to the distribution facility for dispatch. Also, the maximum distribution throughput of the combined operation, at 200 pallets / hour, was also no longer adequate to effectively meet demand.

To solve these issues Walkers set up an engineering project team, and brought on board Swisslog, who would undertake the detailed design work and implement the solution. The concept design that this combined team developed was for a large second high bay warehouse, with space for 13320 pallets, which would be fully integrated with the existing warehouse and the SDRC.

This integration would be achieved by building the new warehouse on the opposite side of the SDRC, from the original warehouse, and then expanding the monorail system to serve both warehouses. The new warehouse, which would be directly linked to the factories, would then feed the monorail from the opposite site of the system to the existing feed, and so utilise the existing picking and dispatch capability and so create a combined operation that could deliver a significant increase in peak throughput, from 200 to 320 pallets / hour.

John Coates, project director for Walkers Snack Foods, explains, "We knew that we would get some important improvements by introducing a new warehouse on the site, and so reducing the need for outside storage, which was costly and inefficient. We also believed that the existing facilities, and in particular the monorail system, had latent distribution capacity that could be utilised to cost effectively deliver a major increase in combined throughput ."

However, while the principle was sound, and management approval for the investment was given the go-ahead, this concept was based on a number of system assumptions, and it was vital that these were fully addressed before detailed design and installation began. As Coates adds, "Significant questions remained as to whether the monorail would simply become too easily grid-locked in practice, when being fed from both sides, and what changes or additions would be needed in equipment, infrastructure, and operational practices to deliver the required throughput of goods."

To answer these questions the joint project team comprising of Walkers, IMI and Swisslog staff, brought in Production Modelling to undertake a detailed process modelling exercise.

Geraint Foulkes, the project manager for Swisslog notes, "Simulation Modelling is a tool that we have used on past projects. However, this was a particularly complex system and so it was important to find the right external partner for this work. In the end, the team looked closely at two suppliers, but we found that Production Modelling's understanding of the situation and complexities involved was far better, and they had the skills and experience to effectively deliver the detailed modelling that was required."

Simulation Strategy

The simulation work had a number of distinct phases. Initially, Production Modelling's team used the ProModel simulation software to model and verify specific elements of the system, at pallet level. In particular, close attention was paid to the configuration of the monorail system.

This model was not only able to answer the key question, that the system could be used as anticipated without any major configuration changes, but also provided important data for the overall project that would have only been found by expensive trial and error. For a start, it had been anticipated that to provide the 320 pallet / hr throughput, up to six more monorail trolleys could be needed. But, the simulation was able to prove that there were already enough trolleys on the existing system to meet this peak demand, which meant a direct project saving of up to £150 000.

Another of the major contributions of the initial simulation exercise was to determine that during the part of the week when the system was not being run at peak throughput, but at a much slower rate of 160 pallets / hr, it was important to park a number of unused trolleys in a new parking area. The empty trolleys when rotated with full ones, jammed the system.

As Coates adds "Overall, the decision to apply simulation technology, and the expertise of Production Modelling, has proved to be a key aspect in the successful delivery of this project."