R-TEK was established in 1991 as a joint venture between the French company J.Reydel and the Kasai Kogyo Co.Ltd. of Japan, subsequent acquisitions have resulted in R-TEK currently being a joint venture between Kasai Kogyo and Visteon Automotive Systems Inc. R-TEK supplies high quality interior automotive components for Nissan Motor Manufacturing UK and Honda UK Manufacturing from its two plants in Washington and Merthyr in the UK. The company currently employs approximately 400 people between the two sites with a turnover of around £38 million in year 2001. R-TEK is run along similar managerial lines to Japanese companies: for instance, R-TEK stresses commitment to product quality and company involvement at all levels. All employees have the same employment guarantees written into their contracts and all are expected to maintain the highest standards of quality in every aspect of the job. Every employee is trained to understand the whole operation, for more effective production.

In 1999 an extension to the Washington plant increased the plant size to almost 16,000sqm, the maximum area available on its current site. This factory is effectively split into three sections: goods inward (stores area), dispatch area (where all goods are shipped to Nissan every twenty minutes or less), and the main production area. The factory is situated only three miles from its major customer, Nissan, and the Washington plant prides itself on its synchronous delivery system. Parts are produced in accordance with Nissan’s production sequence, and shipped by R-TEK’s own drivers, enabling minimum stock levels to be maintained at both the Nissan site and, in theory, internally at R-TEK.

1.1       Performance/strategic objectives

To understand the forces driving R-TEK in its competitive industry it is useful to consider briefly the literature on strategic objectives as they affect performance. At a strategic level, a useful classification of the performance objectives, which any operation might pursue, can be gained from identifying an operation’s stakeholders. Where stakeholders (i) have an interest in the operation, or (ii) may be influenced by, or influence, the operation’s activities. Stakeholders may be internal (for example the operations employees) or external (for example a company’s shareholders). Some external stakeholders have a direct commercial relationship with the organisation, for example the suppliers to the operation and the customers who receive its products and services. (Slack et al, 2001)

The broad objectives which operations must pursue to satisfy stakeholders form the backdrop to all operations decision-making. At an operational level (Slack et al 2001) suggests five basic ‘performance objectives’ which apply to all types of operation:

Quality: providing error-free goods and services, which are ‘fit for their purpose’. At R-TEK all parts are made to the specification set by Nissan and Honda. The products are reliable and blemish-free.

Speed: minimizing the time between a customer ordering goods (or services) and receiving them in full.

Reliability: keeping to delivery promises made to the customer.

Flexibility: being able to change far enough and fast enough to meet customer requirements gives a flexibility advantage to the customer. R-TEK has to be able to cope with product/service flexibility for the introduction of new models and the ability to adjust the number of vehicles manufactured when demand changes.

Price: producing goods (and services) at a cost, which enables them to be priced appropriately for the market while still allowing a return for the organisation.

1.2       Competitive advantage

R-TEK is expected to provide the correct product, at the correct time, with high quality. It competes against companies who are also capable of meeting these order-qualifying criteria (Hill, 1995), and distinguishes itself from the competition by producing to low price (order-winning criteria (Hill, 1995)). According to Christopher (1993) effective logistics management can provide this lower operating cost. The implementation of cost-effective computerised systems is seen as one tool in companies' improving their competitive position by reducing service costs and delivering quality products in a timely manner (Yao and Carlson, 1999).

Within this "low cost" context, this paper reviews an on-going project at the R-TEK Washington plant, which is aimed at acquiring a bar coding system to improve stock accuracy and reduce the costs of held stock within the warehouse. Success in this project will provide one element of the order-winning criterion. Sections 2 to 4 provide a time line through the project as an aid to discuss the issues encountered, Section 5 highlights the main lessons learnt from the project for others to consider when conducting similar projects.

2.         The "Bar Coding" project

In this project a graduate was employed, for two years, through the University of Sunderland. The graduate's role was to work full-time with R-TEK on the project in liaison with academics from both the Computing and Business Schools of the University of Sunderland.

2.1       July '01: The need for a bar-coding system

The project was proposed in July 2001 by the Senior Controller (Logistics). The main focus was to improve the current manual stock control system to increase the level of stock accuracy (which was typically running at 58%). The targets were to have at least 80% accuracy before the end of 2001 and then 98% accuracy before the end of 2002. The stock accuracy figure was calculated using the percentage difference between actual physical stock totals and system totals from a month end stock check. The acquisition of a bar-coding system was expected to reap great benefits to enable this increase in accuracy, and thus generate cost savings in held-stock. The expected annual savings were to be these were to be set against the planned expenditure for the system of approximately £35,0000. The cost of the system was determined by the Senior Controller (Logistics) who took into account the budget available to the warehouse area and made a comparison with some "off the shelf" packages that he identified.

Table 1: Timeline through the Project - at July '01

2.2       October '01: The project start

In October 2001, once the graduate had been recruited, it was decided that to understand the problems of that inaccurate stock figures were causing, and the root cause of these problems, it was necessary to review various areas within the stock handling process. This included both analysing the processes and their operation in practice. For example, one process examined was the physical movement of individual parts through the organisation (from receipt at “goods inward” through to despatch to Nissan). This simple analysis revealed issues that needed addressing, for instance (i) the number of people with uncontrolled access to the warehouse (and stock) and (ii) the large amount of manual data entry being carried out on a daily basis.

2.2.1   The standard process

When the project began, the first six weeks were spent shadowing people in the store to observe daily activities: taking in and checking off deliveries, storing stock, and general maintenance of the stores area. The staff were required to (manually) maintain the paper work connected with new deliveries and returns. This paperwork was passed from stores for input into the company's MRP system after being checked against the delivery schedule and the weekly material planning schedule. If the delivery or corresponding paperwork was incorrect the supplier was informed. When the data was correct the paperwork was passed to accounts for payment.

Each line feeder was required to note what items were needed on the lines and physically move the parts from the store onto the line for the operators. Each item removed from the store should be marked off on an issue sheet (this was removed and a replacement provided at the end of each day). The following day, the data on the issue sheets were used to update the MRP system’s record of the stores stock.

Cycle counts were carried out on a daily basis (at the beginning of day shift before any stock movements were made in the store). The counts were based on each shelving area in the store and arranged so that each item was counted at least once per week. The actual stock figures were then compared, weekly, to the MRP figures to determine the level of stock accuracy, using a spreadsheet. Where the actual and system values differed analysis was performed to determine where the discrepancy arose and why. The system value was then adjusted to reflect reality.

2.2.2   The relationship between existing IT systems and the stock accuracy issue

One of the main problems identified as a result of the initial analysis was the difficulty in effectively using the company's MRP system for stock-related issues. A number of factors influenced this:

§          Stock levels were inaccurate due to poor stock-handling and inaccurate paperwork. The manual paper trail provided numerous opportunities for inaccuracies to be added to the data. For instance, incorrect record keeping when goods arrive, inaccurate logging of moved stock, erroneous data entry into both spreadsheets and the MRP system.

§          The delay in entering/updating the data into the MRP system meant that, at best, the system data was 24 hours behind the current situation.

§          The lack of accurate stock data meant that the MRP system could not be used for the ordering process.

The resultant impact was that stores could experience both problems of “stock-out” and excess stock. Each situation had its own financial repercussions. Where stock was "out" and needed there were two possible outcomes: either (i) failure to supply Nissan as expected (and thus incur financial penalties) or (ii) raising special purchase orders, at an additional cost. Whereas, holding excess stock results in finance being tied-up unnecessarily in goods and thus unavailable for other uses. van den Berg and Zijm (1999) found that the time span between a part arriving in a warehouse and its availability in the computer system is unacceptably long. Companies therefore cover themselves against shortages of material and machine capacity by using safety stocks and safety lead times. By taking such actions inventory is increased and the real problem remains unaddressed. In a Just-In-Time (JIT) environment such problems are not expected to occur. However, Brooks and Wilson (1993) found that the MRP systems and JIT programs adopted by many firms cannot hope to achieve full potential because of inaccurate inventory records.

2.3       January '02: The first findings

In January '02 a formal review was held to discuss the initial findings. Analysis of the cycle counting showed significant problems within the current processes in the warehouse which included: incorrect checking of deliveries, missing paper work, failure to book out stock correctly, and a general underlying cultural (attitude) problem. To overcome these problems the following aims were identified.

2.3.1   Refined aims of the project

Research, and Introduce, Bar Coding Technology. Bar-coding was expected to improve the overall traceability of individual items through the system, (increasing the stock accuracy figure) and also enable a reduction in head count. Moore (2000) suggests that manual collecting and entering part number data for one hundred parts takes approximately ten minutes to record the data and key it in. With bar-coding the same procedure typically takes less than one minute for the same one hundred parts, with an increase in accuracy of the data. It was considered that without bar coding, or some other form of stock-tracing technology, R-TEK would find it difficult, in the long term, to remain competitive.

Improve Staff Morale and Commitment: one of the early problems was staff resistance to new ideas and working methods. Moreover, within stores personnel comments were made to the effect that bar-coding had been tried before, and had failed, … why should it be any different this time? Compounding this was the observation that departments rarely assisted one another on a day-to-day basis, and there was little flexibility within staff roles. Literature records the debilitating effect of employee resistance, lack of departmental co-operation, etc. e.g. (Boehm, 1991; Keil et al, 1998; Moynihan, 1997), therefore, these issues were not unique to R-TEK, but did need to be addressed within this project.

2.3.2   The lack of impact of the findings on the proposed it investment

It had emerged, from comments made by stores staff, that several years previously a bar-coding system had been bought-in, installed, resisted, failed and then abandoned. However, none of the company staff involved in the project had investigated the reasons why, nor reflected on how these might impinge on the current project. Nor had the current project team been made aware of this company history. Moreover, given the acknowledgements of the problems that existed with the MRP system (in terms of its ineffectiveness in supporting inventory management) it might be expected that the linkage between the proposed bar-coding system and the existing MRP system would be a critical requirement. However, the company project-personnel saw the MRP issue as being independent of the stock accuracy problem and the bar-coding solution. The project's Senior Controller was still focused on the goal of acquiring a stand-alone bar-coding system, by April 2002; whereas, the graduate and university team were counselling caution and a need to understand, and improve, the operational (manual) processes before imposing an IT solution. Indeed, the graduate’s operational changes had resulted in increased stock accuracy (and control). The stock accuracy figure improved to 68% in October, and 73% in November. Although there was a significant drop in December to 55%, this was related to the launch of the new Nissan Primera and the change over from old parts to new. 

Table 2: Timeline through the Project - at January 2002

2.4       February '02: Literature review and comparisons

In February 2002, to identify the required features of suitable systems for IT evaluation purposes the project undertook literature reviews within the areas of stock accuracy e.g. (Brooks and Wilson, 1993; Lipschitz, 2001; Piasecki, 2000b), inventory control and management, e.g. (Piasecki, 2000a; van den Berg and Zijm, 1999; Schreibfeder, 1997a,b and 2002; Yao and Carlson, 1999), and data capture/bar-coding, e.g. (Lindau and Lumsden, 1999; Manthou and Vlachopoulou, 2001). This was complemented by an investigation of the current commercial IT packages, and case studies of companies that had implemented such systems successfully. Analysis of the results indicated that bar coding was indeed the way forward for the warehouse, and should eventually lead to the introduction of a fully automated system throughout the factory. Schreibfeder (1997a,b), and Brooks and Wilson (1993) contend that without accurate computerised stock level data effective inventory management is impossible since the benefits from a good inventory management system are only gained when stock balances have consistently a 95% minimum accuracy. Such a consistent accuracy is impossible to achieve manually, at an acceptable cost. Moreover, a fully automated system should enable full traceability of stock, a faster response time and streamlining of the operation; thus providing significant cost reductions to support the company's competitiveness.

2.5       April '02: A period of uncertainty.

As Easter 2002 approached the project progress slowed as the company had begun to appreciate difficulty of integrating a bar-coding system with the current MRP system. This was now seen as a pivotal part of the project. Moreover, it became apparent that other departments had been independently investigating the replacement of the current MRP system without any awareness of the bar-coding study. Thus, time was spent informing each other of respective needs and preferred options. Concurrently, the Senior Controller (Logistics) the project resigned. The uncertainty regarding a replacement MRP (incorporating inventory management) and the loss of the project champion destabilised the project. However, after an emergency review, the Deputy Managing Director (DMD) agreed to lead the project.

Subsequent discussions with the DMD established that this project was of a strategic nature. The need for increased stock accuracy stemmed from a need to reduce stock held - and ultimately this was linked to a strategic aim. This aim was to ensure that the company could respond to the introduction of new models at Nissan (requiring new product parts) and, ultimately, win orders from other organisations (with their associated product parts) without the need to invest in costly expansion of the plant and its stores area. The DMD acknowledged that, for the project to be successful, its profile had to be raised significantly so all staff were aware of its importance to the future success of R-TEK. Any bar-coding (or inventory management) system that was to be introduced needed to be appreciated for its strategic value, not just as an operational tool. Table 3 highlights some of the changes to the project and the evaluation of its need for an IT system over this period.

Table 3: Timeline through the Project - at April 2002

N.B. During this period of uncertainty the rigorous application of the newly implemented stores procedures lapsed, with a resultant "dip" in that month's stock accuracy (See Chart 1).

3.         April to December 2002: A strategic project

Thus, after six months the project was re-evaluated and the experience of other companies and academic literature were intensively reviewed to provide a more effective focus. From this investigation two factors emerged as being vital to the success of the project: the introduction of documented procedures and effective cycle counting.

3.1       Procedure documentation

Before any progress could be made in developing an appropriate and detailed specification for evaluating potential IT solutions the processes that were affecting stock accuracy throughout the organisation had to be clearly defined (Manthou and Vlachopoulou, 2001). In April 2002 the warehouse was the only department that did not have any form of standard operations documented for employees to follow and, although job descriptions did exist, the roles had altered since their introduction. Most warehouse employees had devised their own work practices and, whilst the general principles of stock handling remained constant, there was a serious lack of standardisation. Therefore, new job descriptions were issued and standard operations were introduced to match the current, refined, procedures that had been developed over the first six month of the project. These new procedures facilitated accurate data entry and provided an explicit audit trail linking data entry/stock changes to individual employees.

3.2       Cycle counting

Analysing R-TEK's situation against the results of the literature review it was clear that cycle counting was going to play a vital part in the continuous improvements to stock accuracy (Piasecki, 2000b). One dedicated cycle counter was assigned to count parts in the warehouse on a geographical basis split across five working days. However, for the cycle counting to be successful the full dedication of all employees involved and consistency was vital to the entire project.

3.3       Changing the culture

Literature highlights the importance of making employees feel responsible and included in their organisation, e.g. (Schreibfeder, 1997a). Therefore, at R-TEK, in addition to the formalised procedures regular meetings were instituted, with all stores personnel. These meetings informed of changes and future plans, in addition staff were free to raise any concerns and these were explicitly dealt with and tracked. Managers also ensured staff were acknowledged for improvements made and new ideas suggested: this was a new concept within the company. As a result, the (initially resistant) staff felt valued as part of a successful team that was pivotal in improving the company's competitiveness.

3.4       Stock accuracy results over the project lifespan.

Over the lifespan of the project the stock accuracy figures had an upward trend from the starting value of 68%, although there were isolated "drops". These results can be seen in Chart 1 below. In April 2002 during the project's "period of uncertainty" the figure dropped to 57%. In July 2002 there was a dip to 68%. Detailed analysis revealed that (i) the original formula used had an inaccuracy in it, which, although normally insignificant, had been partly responsible for the decline, (ii) a poor quality count had taken place using untrained employees (resulting in duplicated data entry and an overstocked stores). Thereafter, greater emphasis was put onto the (i) analysis of the cycle counts, (ii) use of the new procedures. The stock accuracy figure continued to rise steadily. In November 2002 the new Nissan Micra was launched and this was seen as the first major test of the improvements. In December 2001 the launch of the new Nissan Primera had resulted in a drop in stock accuracy of 18% (as shown in Chart 1) and an increase in stock holding in the warehouse, whereas the stock accuracy for November 2002 dropped by only 3% to 82%. This confirmed that the new manual system was robust enough to cope with major perturbations, not only within the warehouse but also throughout the entire factory. Moreover, in December 2002 the stock accuracy figure was 90%.

Since July 2002 there have been few significant changes to the manual system, and the effectiveness of it, are reflected in the subsequent results. Chart 1 demonstrates that the manual system can keep stock accuracy at above 80% (the initial target set for December 2001, but not achieved until October 2002) and routinely, with effective procedures, in the range of 85-90%. However, the final target set for accuracy of 98% or above has never been met; it is unlikely that any manual system could ever achieve such precision (Schreibfeder, 1997a,b; Brooks and Wilson, 1993). It is here, therefore, in raising accuracy to that final level that a supportive IT system can play a part.

Chart 1: Stock Accuracy from October 2001 to June 2003.

4.         Preparing for inventory management IT systems support.

Throughout the time period over which the manual processes, and stock accuracy, were increased a second activity was progressing: this was the evaluation of potential support systems. This evaluation took two forms: review of literature (academic, trade and "sales") and sites visits for inventory systems. The literature review (as outlined in Section 2.4) enabled the graduate to refine her understanding of the features that would be beneficial in such systems, and that should be available. The evaluation of sales material helped her to identify realistic cost parameters for such systems (including hardware, software, support, training) and provided evidence to substantiate the case for increasing the budget allocation. Table 4 shows how the projected cost of the system has risen steadily from the initial estimate of £35,000 to a more realistic figure of £100,000.

Table 4: Timeline through the Project - at October 2003

Five site visits were made to companies operating bar-coding and inventory management systems. These visits highlighted the actual benefits of the technology as well as, importantly, showing how easy it was to undermine its effectiveness. During the site visits the users of the systems spoke freely of their experiences, and even where they had a positive view of the technology, they all indicated problems and the importance of having in place a sound (manual) process. A second set of site visits were held with potential vendors visiting R-TEK Washington and (i) demonstrating their standard products and (ii) identifying how they would adapt these to the existing R-TEK processes. As the project progressed the vendors were also required to identify their solutions for integrating with the existing MRP system, and facilitating the move to a new MRP system.

The combination of visits (off and on site) greatly increased the confidence of the company-project team in defining their specification and evaluating potential vendors. However, without the development of rigorous, proven, manual processes such confidence in specifying the requirements and evaluating potential solutions could never have been achieved. The current position within the project is that two preferred vendors have been identified and their products are under detailed review. The implementation date has now been put on hold as the accuracy and control within the manual process has substantially increased the company's ability to manage and control stock levels and deal with unexpected events. A detailed cost-benefit analysis is planned to justify the implementation of the bar code system and how this will provide an overall benefit to the company and the manner in which costs-savings in head-count and reduced inventory will outweigh (over time) the IT investment.

5.         Conclusion

This case study has demonstrated that many factors can impact on IT evaluation and procurement. One of the most critical however, is a sound basis from which to make judgements for the need for, and the justification of, a system. In this case study could only be achieved by undertaking business process improvement before committing to an IT solution. Other significant factors that were experienced within the project, which resulted in it being considered a success were:

§          The identification and involvement of a project champion who had authority,

§          Communication between, and involvement of, different stakeholders,

§          The rationalisation and improvement of processes before IT acquisition resulting in cost savings.

§          Development of a strategic view of what the system will support

§          Evaluation of the IT system need based on realistic costs and justified against realistic savings (e.g. stock-held, head count).

§          Growing awareness of the range of issues that impact on such decision making (e.g. integration with existing systems).

6.         Acknowledgements

This project was partially supported through a technology-transfer grant from the UK Government agency, One North East, as part of the "Graduate Retention in the Automotive Sector Programme (GRASP)".

References

• Boehm, B (1991) Software Risk Management: Principle and Practices. IEEE Software, 8(1), pp32-41.
• Brooks, R. and Wilson, L (1993). Inventory Record Accuracy Unleashing the Power of Cycle Counting. Oliver Wight Publications Inc, Vermont
• Christopher, M. (1993). Logistics and Supply Chain Management. Pitman Publishing, London.
• Hill, T. (1995) Manufacturing Strategy. Text and Cases. Macmillan Business, Basingstoke.
• Keil, M., Cule, P.E., Lyytinen, K., and Schmidt, R. (1998) A Framework for Identifying Software Project Risk. Communications of the ACM, 41(11), November 1998, pp76-83.
• Lindau, R. and Lumsden, K. (1999). The Use of Automatic Data Capture Systems in Inventory Management. International Journal of Production Economics (59) pp. 159 – 167.
• Lipschitz, D. (2001). The Requirements of Stock Accuracy. Orbital Decisions.
• www.stockaccuracy.co.za. Accessed 10/04/2002
• Manthou, V. and Vlachopoulou, M. (2001). Bar-code Technology for Inventory and Marketing Management Systems: A model for its Development and Implementation. International Journal of Production Economics (71) pp. 157 – 16.
• Moore, B. (2000). Introduction to AIDC Technologies. IDAT (Intelligent Data Acquisition Technologies). www.idat.com/tech.html. Accessed 20/02/02
• Moynihan, T. (1997) How Experienced Project Managers Assess Risk. IEEE Software, 14(3), May/June 1997, pp35-41.
• Piasecki, D (2000a). Guide to Inventory Accuracy. Inventory Operations Consulting. www.inventoryops.com/guide_inv_acc.htm. Accessed 10/04/2002.
• Piasecki, D. (2000b). Cycle Counting and Physical Inventories. Inventory Operations Consulting. www.inventoryops.com/cycle_counting_and_physical_inventories.htm.
• Accessed 10/04/2002.
• Schreibfeder, J. (1997a). Do Your Employees Understand the True Cost of Lost Material? Effective Inventory Management Inc.
• www.effectiveinventory.com/article7.html. Accessed 20/03/2002
• Schreibfeder, J. (1997b). Cycle Counting Can Eliminate Your Annual Physical Inventory. Effective Inventory Management Inc.
•  www.effectiveinventory.com/article9.html Accessed 20/03/2002
• Schreibfeder, J. (2002). Achieving Cycle Counting Success. Effective Inventory Management Inc. www.effectiveinventory.com/article48.html. Accessed 20/03/2002
• Slack, N. Chambers, S. Johnston, R. (2001). Operations Management. Pearson Education Limited, Essex.
• van den Berg .J.P. and Zijm W.H.M. (1999). Models for Warehouse Management: Classification and Examples. International Journal of Production Economics (59) pp. 519 – 528.
• Yao, A.C., and Carlson, J.G. (1999). The Impact of Real-time Data Communication on Inventory Management. International Journal of Production Economics (59) pp. 213 – 219.

EJISE is published by Academic Conferences Limited
Curtis Farm, Kidmore End, Nr Reading RG4 9AY, England
Tel: +44 (0)1189 724148, Fax: +44 (0)1189 724691, Email: info@ejise.com

Send mail to info@academic-conferences.com with questions or comments about this web site.
Copyright © 2002-2006 Electronic Journal of Information Systems Evaluation
Last modified: September 29, 2005
ISSN 1566-6379