2 PHARMACEUTICAL ENGINEERING • NOVEMBER/DECEMBER 2001

Flow Path Management

manufacturing excellence. The following measurable improve-ments have taken place throughout the operation without thesignificant addition of capital assets or human resources:

• Units shipped are up approximately 23% over a two-yearperiod.

• Cycle time has been cut by 50% for high volume, high valueproducts.

• Inventory is $10 million lower for critical materials.

• On-time delivery performance has been dramatically im-proved and is now nearly perfect.

This article summarizes the philosophies, organizationalchanges, and information systems that were implemented inorder to achieve manufacturing excellence. It concentrates onthe planning and scheduling improvements and flow pathmanagement techniques that have cut cycle time and stream-lined product flow. The article describes the following twophases:

• Phase One: Developing the Planning and Scheduling Infra-structure

• Phase Two: Flow Path Management to Improve Perfor-mance and Cut Cycle Time

Plant MissionDuPont’s 370 employees produce a total of 150 product Stock-Keeping Units (SKUs) ranging from tablets and capsules tosyrups. A wide variety of processing technology is employed atthe plant including direct compression, wet granulation, rollercompaction, fluid bed drying, tray drying, tablet compression,aqueous film coating, and encapsulation. The site packages alarge number of bottle configurations and also blisters. Thefacility was built in 1963 and has grown in a series of expan-sions to a total of 142,000 square feet.

The process flows for oral solids at Garden City are typicalfor the industry. Raw materials are received, sampled, andtested. The materials are then weighed and blended intopowder, which is either compressed into tablets or encapsu-lated. Most tablets are coated and tested prior to packaging.Finished products are packed and tested prior to final ship-ment.

The plant has a two-fold mission:

1. Partner with R&D to rapidly scale-up and launch newpharmaceutical products. The plant is involved in 10-12

product development projects per year. Some of these prod-ucts are later transferred to a sister location in Manati,Puerto Rico.

2. Supply existing products to the market in a high-quality,economical fashion. The plant produces about 40 differentcommercial products and 150 SKUs.

The business challenges are clearly different for each part ofthe mission. New products have relatively unpredictable de-mand curves and require care and feeding as the organizationclimbs its learning curve. High volume products, on the otherhand, must be produced with consistent on-time delivery andefficient operational costs.

Starting Point: The Need for ImprovementsThree years ago, the site operated in a fashion that wassomewhat typical for the pharmaceutical industry:

• The organizational structure was functionally oriented.Each department used stand-alone systems and perfor-mance metrics.

• Individual departments maintained “hot lists” based ontheir knowledge of required customer ship dates.

• There was limited product flow. Material typically spent90% or more of the time waiting for the next operation. Thisresulted in long product cycle times, excessive inventory,and costly material storage and handling.

• Systems were not integrated. The ERP system and the MRPmodule within ERP was used only to manage inventory anddo financial reporting. Inventory transactions were one tothree weeks behind.

• Planning and scheduling activities were shortsighted andmanual-intensive. The planning horizon was one month atmost. Schedules were manually maintained using spread-sheets - hard copies were circulated weekly, often withhandwritten notes. Planners tracked work center activitiesusing telephone calls and safety shoe leather. Despite theextra effort, the schedules often had to be re-issued severaltimes each week to track changes. Moreover, several impor-tant functions such as Quality Control and Quality Assur-ance were excluded from the initial scheduling processbecause of limited planning staff.

• Strategic capacity planning for capital and human re-sources was not formalized.

Table A. Business objectives for each Focused Factory and Flow Path.

Factory Resouce and Equipment Priority Goal Detailed Scheduling Method

1. Development, validation, and Top Priority Speed-to-market for new products Manual review and controllaunch

2. High volume, high cost Dedicated equipment where Low inventory, fast cycle times Pull Scheduling using CONWIPproducts possible, high priority allocation of

people and equipment

3. Low volume, low cost Shared equipment, lower priority No stock-outs Push ERP/MRP scheduling withproducts for resource allocation generous lead times - use

finished product safety stock as abuffer

NOVEMBER/DECEMBER 2001 • PHARMACEUTICAL ENGINEERING 3

Flow Path Management

To address these shortfalls, the objective was defined to im-prove planning and manufacturing performance. The im-provements were performed in two phases: Phase One builtthe information systems and planning and scheduling infra-structure to support improved performance. Phase Two builton this foundation by splitting the operation into severaldistinct product Flow Paths, and then redesigning the organi-zation structure, performance metrics, equipment capabili-ties, and planning and scheduling tools to meet the needs ofeach flow path. Detailed performance objectives for the orga-nization also were introduced as part of Phase Two.

The following sections describe each phase in detail.

Phase One of Change: Developing the Planningand Scheduling Infrastructure

The change process began by strengthening the informationsystems and planning and scheduling tools. This positionedthe information systems to function as a true ERP system,rather than just an inventory management and financialreporting system.

An operation management system was built, one piece at atime, starting with the ERP backbone. Figure 1 shows themajor planning and scheduling activities improved in PhaseOne. Figure 2 shows the activities improved in Phase Two.

ERP Data Collection and AnalysisA basic first step was to populate the ERP system withaccurate routings and work center data. This involved a jointeffort between the site Planning organization and the Manu-

facturing and Packaging work groups. The pertinent datafields were populated and validated for accuracy over a six-month period. The ERP system was designed with a tremen-dous amount of detailed data - this investment paid hand-somely in Phase Two.

Having the ERP data in place, the goal was to consolidateoperating and decision making information within ERP. Thegoal was to track inventory and production transactions real-time.

Inventory and Production TrackingOur inventory tracking practices were state of the art for 1965:manual keypunching of inventory transactions into the ERPsystem by a few trained individuals. This caused long delays,large errors, excessive inventory, and, as a result, materialshortages and line shutdowns.

How would up to 450 inventory movement, consumption,and production reporting transactions per day be performedwhile maintaining up-to-the-minute accuracy? Commerciallyavailable technology would allow barcode scanning of materi-als as actual physical activities took place, followed by adatabase upload to the ERP system from the handheld scan-ning devices. But this would delay ERP updates by the fre-quency of uploads. Not wanting to be limited by aged data, aproprietary technology called Radio-frequency Order Pickingand Inventory Control System (ROPICS) was selected. Thissystem uses wireless handheld barcode scanning devices toinstantly transmit data to and from the ERP system.

Everyone who touches materials was trained to perform

Figure 2. Planning and scheduling system components improved in Phase Two.

4 PHARMACEUTICAL ENGINEERING • NOVEMBER/DECEMBER 2001

Flow Path Management

Figure 4. System architecture for flow path management metrics. Bar code scanners feed information into the ERP system and data warehouse. These are displayed for allemployees via the company intranet site.

ROPICS transactions as part of any given operation theyperform. As a way to ensure ROPICS transaction discipline,cycle count accuracy was measured and actively managed to beat least 95% on a consistent basis. All “misses” are investigatedand corrective actions implemented.

What was the result? Inventory accuracy improved from70% to 95+% within six months of implementing the newapproach. This ensures cGMP compliance when accounting forcritical materials. In addition, very significant quantities ofraw materials have been removed from inventory, reducingworking capital by more than $10 million and cutting cycletimes for high volume products by 50%. This is directly relatedto inventory record accuracy within ERP - when you knowexactly what you have at any given point in time, you don’tneed any “padding.”

Our manufacturing and information technology staff hasrecently leveraged the ROPICS technology for managing theshop floor weigh-up and charge-in processes. This has greatlyenhanced the ability to perform these critical operations in anerror-free fashion by eliminating manual calculations and byguiding batch formulation real-time using a validated system.New ways will continue to be explored to use this technology forshop floor operations.

Since improving inventory accuracy, the materials require-ment plan from ERP has been made much more meaningful byhaving the rough-cut schedule and real time inventory bothresident in ERP. We just don’t run out of materials anymore.

Shop Floor Control, Sequencing, and SchedulingCustom reports were developed to show a weekly rough-cutschedule and equipment capacity utilization for each of the 75Manufacturing and Packaging work centers - all based on ERPshop orders. These reports replaced the spreadsheets and arenow the primary means for communicating the rough-cutschedule.

Quality Control and Quality Assurance were initially leftout of the ERP scope. This was a mistake. Since the initialproject, a program was launched and completed to add thesework centers to the routings so that they are integrated withthe Manufacturing and Packaging operations. The addition ofQA and QC to the implementation scope was critical forreducing cycle times, since products interface with these workcenters up to five times during each cycle.

Capacity/Facility PlanningWith the tactical elements in place, we were able to lengthenthe planning horizon to the strategic time frame. Equipmentcapacity utilization can now be predicted for each work centerthroughout the forecasting horizon (two years in this case).

Figure 3. Volumes for each Garden City Flow Path. This graph shows shipments byflow path. It includes historical (prior to 2001) and forecasted (2001 - 2003)volumes.

NOVEMBER/DECEMBER 2001 • PHARMACEUTICAL ENGINEERING 5

Flow Path Management

This is done using the ERP capacity planning modules withlittle customization. The same techniques are used to calculatelabor requirements for each work center, knowing only thelabor rates and product demand stored in ERP.

The Directors of Engineering, Manufacturing and Packag-ing, Technical Operations, Human Resources, Finance, Infor-mation Resources, and Planning now meet monthly to reviewthe capacity plan for the upcoming two years. All capitalequipment and staffing plans are formulated from these meet-ings and incorporated into the budgetary process. Havingsufficient but not excessive equipment capacity and laboravailable at the proper times makes the best use of thecompany’s cash and provides the necessary conditions forreduced cycle time. The site’s financial budgets and long-rangeoperating plan also are developed from the ERP database.

Phase II of Change: Flow Path ManagementDefining Flow Paths using Pareto AnalysisPerformance soon peaked following these initial successes.Relying on ERP and the traditional performance metrics wasreducing efficiency, increasing inventory, and lengtheningcycle time.7

Given the duality of the plant’s mission statement, thecomplexity of the manufacturing operation, and the limitedequipment flexibility, we needed to focus on specific productgroups and tailor the management systems for each productgroup.

Flow Path Management is a management technique thatorganizes manufacturing systems into process-based flowpaths.8 These flow paths simplify planning and scheduling,support organization structures aligned to process flow, andenable cycle time reductions and other performance improve-ments. Management can tailor the business processes in eachof the following four areas:

1. Performance Metrics

2. Organization Structure and Development

3. Planning and Scheduling

4. Process Control and Equipment Flexibility

To identify the major flow paths, Pareto analysis was used. Theanalysis showed that the bulk of inventory dollars were in-vested in two products. We decided to focus on reducing cycletimes for these two high volume, high value products. Theactual Garden City flow paths along with their associatedvolumes are shown in Figure 3.

The operation was divided into three distinct factorieswithin the overall operation, called “factories within the fac-tory.”9 Development products were assigned to the first fac-tory. The second factory, for high volume products, held two

flow paths: one for each of the two high volume products. Theremaining low volume products were assigned to the thirdfactory. The strategy to accomplish the entire mission involvedthe scheduling rules shown in Table A.

After defining the major flow paths, the business processeswere tailored within each flow path to best fit the needs of thatbusiness.

Performance Metrics and Feedback InformationTo track progress and to motivate and reward improvement, aset of performance measures for each flow path was designedand implemented.

The vision was to automatically generate these metrics ona daily basis and to give individual work center owners theability to view them on that frequency. To implement thisvision, a data warehouse was developed containing all the ERPdata and inventory transactions described above. We call thisour Cycle Time Information System (CTIS). Using CTIS, workcenter owners are able to quickly visualize, on a daily basis,their own area’s performance for each flow path and make anynecessary adjustments. Figure 4 shows how the data is auto-matically transferred from the ROPICS guns into the datawarehouse, and then analyzed on our intranet web server fordisplaying the performance for a given flow path. Every personat the plant (and in the company at other locations) can viewthe metrics using the corporate intranet site. Figure 5 showsan example of CTIS output, displaying the location and on-time status of open orders in a flow path.

These metrics are reviewed at the daily operations meetingto make operating decisions and monitor status for each flowpath. The purpose of this meeting is to establish a common

““ ““The Cycle Time Information System is usedas an early warning system

so that bottlenecks are identified and resolved quickly.

Figure 5. Measuring flow path health to drive operational excellence. This graphshows how many orders are waiting at each operation along the flow path.Employees can see if the orders are on time using the legend on the right, and candouble click on the graph to see details for late or behind schedule orders. Thegraph is automatically updated daily, and is published for all employees on theintranet site. (Similar graphs show cycle time vs. goal and inventory vs. goal).

6 PHARMACEUTICAL ENGINEERING • NOVEMBER/DECEMBER 2001

Flow Path Management

Figure 6. Pull scheduling using Constant Work in Progress (CONWIP). This diagramillustrates how CONWIP signals the first operation to start more Work In Progressinventory (WIP) when inventory levels drop below the target. Note that multipleproducts can be scheduled, and that equipment can be shared across products.

understanding of the three factories’ status each day and toprovide a vehicle for communication among the work centerowners. Hundreds of active shop orders are managed (byexception) cross-functionally in a daily 15- minute meeting.Issues are identified early and resolved.

CTIS is used as an early warning system so that flow pathbottlenecks are identified and resolved quickly. Three FlowPath Metrics are monitored daily:

• Health of the Flow Path: Displays the location of openorders within each flow path (dynamic bottleneck identifi-cation) and whether the orders are on time, behind sched-ule, or late. Figure 5 shows this graph.

• Cycle Time vs. Goal: Measures elapsed time from rawmaterial (active ingredient) receipt to shipping of finishedproduct for each flow path. Provides overall measure ofoperational effectiveness and motivates cycle time reduc-tion efforts.10-14

• Inventory vs. Goal: Determines when to pull work into theflow path. This pull scheduling tool controls work in progressinventory and cycle time.15

Distribution of the information through CTIS provides a con-stant source of performance feedback. Using CTIS, we are ableto break flow path cycle time into its individual componentsand encourage continuous improvement from each work cen-ter owner.

Site operating objectives are built around flow path perfor-mance for each of the three flow paths. Stretch goals are setwith aggressive improvement. Every functional group has itsown objectives that are based on the whole organization’sobjectives. All 370 people in the organization can articulatewhat role they play in achieving these objectives, since theobjectives are an element of each person’s performance ap-praisal.

A portion of each person’s compensation is determined bythe success of the organization in meeting site operationalobjectives. The performance management process is carefullymonitored at the highest levels of the site organization toensure that the compensation process is used effectively todrive manufacturing excellence. The site leadership team isresponsible to ensure that each objective is approached in asynergistic fashion. This reinforces the site objectives andmotivates individual contributions and team performance.16,17

Organization Design: Aligning People and SkillsThe plant took several steps so that the organization reachesstretch objectives:

• Additional training was offered to help employees expandtheir skills. The training included an on-site workshop inFactory Physics techniques to expose everyone to the oppor-tunities for cycle time reductions and other improvements.Factory Physics is a systematic description of the underly-ing behavior of manufacturing systems. These analysistechniques are used to identify opportunities for improve-ment and target specific improvement projects.3 The totalsite training budget has been approximately $250,000 peryear over the last three years. In addition, job-specifictraining is managed as part of the normal course of busi-ness.

• Strategic partners were hired to speed implementation andcomplement the skills of internal staff members. DuPontPharmaceuticals retained the services of SAK Logistics tosupport the implementation of the Cycle Time InformationSystem and the cycle time reduction projects.

• Additional people were dedicated to process improvementand cost savings. As part of a corporate Six Sigma processimprovement initiative, two DuPont Pharmaceuticals em-ployees were appointed to serve as full-time Six SigmaBlack Belts.

• People were assigned to specific flow paths. By focusing onthe needs of just one flow path, the employees can customizetheir work and improve cycle time through their area.

• People within a flow path were cross-trained so that theyare able to move to the flow path bottleneck and relieve thecongestion. As an example, suppose the tablet coatingoperation experienced downtime causing uncoated tabletcores to accumulate in front of that operation. The flow pathhealth metrics would quickly highlight the bottleneck.Cross training allows people from the weigh-up operation tomove to coating and increase production of coated tabletsuntil the bottleneck is relieved.

• Staff groups were redeployed to support flow path success.Technical and QA resources were made available to the

Figure 7. Reducing changeover times using SMED techniques. This graph shows thechangeover time reductions achieved at one packaging line using Single MinuteExchange of Dies (SMED) techniques.

NOVEMBER/DECEMBER 2001 • PHARMACEUTICAL ENGINEERING 7

Flow Path Management

operating areas 24/7 for consulting and problem solving.Material is not permitted to move from one operation to thenext without complete satisfaction that all work was per-formed correctly. Focusing on problems as they occur in-creases visibility, drives more participation in the problemsolving process, and creates more ownership for the qualityof the product. This minimizes cycle time variability.

Pull Scheduling to Cut Cycle Time and InventoryWith the metrics and organization in place, earlier planningand scheduling improvements were expanded to reach betterperformance. One primary obstacle to overcome was the weak-nesses of Materials Requirements Planning (MRP) as a sched-uling tool. While MRP works well in plants with limitedcomplexity and with low utilizations, it begins to fall apart asreal-world conditions occur.18 For example, MRP is not sensi-tive to capacity constraints. As a result, plants using MRPsystems have an incentive to “pad” their lead times just in casevariability causes a production problem. These padded leadtimes increase inventory and cycle times unnecessarily. Inaddition, since MRP is a “push” system, it will not slow downmaterial releases if there is a production problem in thefactory. This causes inventory to build needlessly and in-creases cycle times further.3

To overcome these issues, a pull scheduling system wasimplemented for high volume flow paths. A Constant Work in

Progress (CONWIP) system15 was adopted - Figure 6. CONWIPworks by setting an inventory target for each flow path. Wheninventory drops below the target, the system sends a signal tothe first operation to send more work into the flow path. Thispull signal works to always maintain a constant amount ofwork in progress in the flow path. Planners and work centerowners use the pull signal contemporaneously with flow pathhealth information to schedule each factory flow path andmonitor daily performance. Using current, accurate, focusedflow path data allows optimum communication and coordina-tion.

The CONWIP system has several advantages over MRP.First, it allows inventory and cycle time targets to be reached.Second, it eliminates the motivation to pad MRP lead times.Finally, it requires significantly less data than MRP since onlya single number is needed — the total inventory for the flowpath each day.

CONWIP also offers several advantages over other pullscheduling methods such as Kanban cards. First, CONWIPworks well even if many low volume products are produced onthe same flow path; the pull signal authorizes the release of thenext order for the flow path, regardless of the specific SKU.Second, CONWIP provides a simple way to move material tothe bottleneck for the flow path. As an example, if the tabletcoating operation experienced downtime, uncoated tablet coreswould accumulate in front of that operation without anychanges in the CONWIP target.

Improving Equipment Flexibility and Reliability toSupport Faster Cycle TimeWith pull scheduling, inventory levels can be lowered to anylevel. However, care must be taken to avoid setting inventorylevels so low that bottleneck equipment starves for work. Inshort, enough inventory must be maintained to handle ex-pected equipment outages, changeovers, and between-lot clean-ing. The more variability in equipment uptime or the longerthe changeover times, the more inventory that is required.

As part of the change process, several work centers wereidentified with long outages and/or long changeover times.Teams were formed to improve this equipment. For example,one team was formed at the high speed packaging line toreduce changeover times. This team included operators, me-chanics, QA technicians, and support personnel from all threeshifts. As shown in Figure 7, the team was able to cut changeovertimes by more than 50% in 90 days using Single MinuteExchange of Dies (SMED) techniques.19 By reducing thechangeover times, the equipment utilization dropped, reduc-ing cycle time and facilitating reductions in campaign sizes,operating costs, and inventory.

Results and Lessons LearnedThe Garden City organization has validated tens of millions ofdollars in working capital reductions by improving the wayinventory is managed. Shipments have reached record highlevels, and cycle times continue to set new monthly records -Figure 8. We are confident that this is just the beginning of thebenefits that will be realized.

The key lessons learned include:

• ensure that information systems are timely and accurate

• identify product flow paths based on business needs

Figure 8. The dollar value of cycle time reductions. This graph shows the cycle timereductions achieved for the high volume flow path. The upper graph shows howcycle times were reduced from 142 days to 70 days. The lower graph calculateshow this reduction reduced working capital (inventory dollars on the balance sheet)by roughly $10 million.

8 PHARMACEUTICAL ENGINEERING • NOVEMBER/DECEMBER 2001

Flow Path Management

• establish flow path performance metrics and metric owner-ship to motivate improvements

• staff each flow path and train its employees so they worktogether to reach stretch objectives

• utilize external strategic partners to accelerate implemen-tation and complement internal skills

• use pull scheduling to overcome the weaknesses of MRP

• continued improvement requires faster changeovers andsmaller campaign sizes

References1. Skinner, W. Manufacturing: The Formidable Competitive

Weapon. Wiley, New York, 1985.

2. Dertouzos, M.L., R.K. Lester, and R.M. Solow. Made inAmerica: Regaining the Productive Edge. MIT Press, Cam-bridge, MA.,1989.

3. Hopp, W.J.; M.L. Spearman, M.L., Factory Physics, Foun-dations of Manufacturing Management, Second Edition,McGraw Hill, New York, 2001.

4. Boudette, N., “Europe’s SAP Scrambles to Stem Big Glitches- Software Giant to Tighten Its Watch After Snafus atWhirlpool, Hershey,” The Wall Street Journal, Nov 4, 1992.

5. Stedman, C., “Survey: ERP costs more than measurableROI,” Computerworld, April 5, 1999.

6. Bartholomew, D. “Lean vs. ERP,” Industry Week, July 19,1999, 24-30.

7. Peterson, Rein, and Silver, Edward A. Decision Systems forInventory Management and Production Planning. JohnWiley and Sons, New York, 1979.

8. Knight, Thomas P., Inventory Reduction in a Large JobShop, Masters Thesis, MIT, Cambridge, MA 1992.

9. Schonberger, Richard J., World Class Manufacturing: TheLessons of Simplicity Applied, MacMillan, New York, 1986.

10. Schmenner, Roger W. “The Merit of Making Things Fast,”Sloan Management Review, Vol. 30, No. 1, pp. 11-17, Fall1988.

11. Karmarkar, Uday. “Getting Control of Just-in-Time,”Harvard Business Review, Vol. 1, No. 5, pp. 122-131, Sept-Oct 1989

12. Knight, Thomas P., Kramer, Mark A., and Rosenfield,Donald B., “Measures to Drive Flexibility and ReliabilityImprovements,” Manufacturing Measurements - 21st Cen-tury Metrics Conference Proceedings, pp. 88-98, LehighUniversity, Bethlehem, PA, 1992.

13. Maslaton, Rafi., “Creating Manufacturing ExcellenceThrough Cycle Time Reduction.” Pharmaceutical Engi-neering, Vol. 20, No. 4, July/August 2000, pp. 38-46.

14. Suri, Rajan. Quick Response Manufacturing: A Company-Wide Approach to Reducing Lead Times. Productivity Press,Portland, 1998.

15. Spearman, M.L., Woodruff, D.L., and Hopp, W.J., “CONWIP:A Pull Alternative to Kanban,” International Journal ofProduction Research, Vol. 28, No. 5, pp. 879-894, 1989.

16. Gerecke, Glenn A., Team-Based Compensation: A ModernManagement Dilemma, Masters Thesis, WPI, Worcester,MA 1998.

17. Miller, Stanley S., “Make Your Plant Manager’s Job Man-ageable,” Harvard Business Review, pp. 69-74, Jan-Feb1983.

18. Boudette, N., “Europe’s SAP Scrambles to Stem Big Glitches- Software Giant to Tighten Its Watch After Snafus atWhirlpool, Hershey,” The Wall Street Journal, Nov 4, 1999.

19. Shingo, Shigeo. A Revolution in Manufacturing: The SMEDSystem. Productivity Press, Portland, OR, 1985.

About the AuthorsGlenn Gerecke has approximately 20 years of work experi-ence in the chemical processing and pharmaceutical indus-tries. His career is characterized by leadership in the areas oforganizational design and development, invention and imple-mentation of systems, technological innovation, operationalquality, and project management. Gerecke is currently VicePresident and Site Director, DuPont Pharmaceuticals Com-pany, Garden City, New York. (DuPont was acquired byBristol-Myers Squibb Company on October 1, 2001.) He isresponsible for all aspects of site management and leadership.In this position, Gerecke has built an exceptionally strong andextremely motivated organization, has driven remarkableimprovements in first-time-through quality and cycle timereduction, and has implemented leading edge operationalsystems. He has Masters’ degrees in business administrationand management from Framingham State College and Worces-ter Polytechnic Institute, respectively. He holds a BS in chemi-cal engineering from Worcester Polytechnic Institute.

Bristol-Myers Squibb Company (Formerly Dupont Phar-maceuticals Company), 1000 Stewart Ave., Garden City, NY11530. (516) 832-2002.

Tom Knight is a Partner with SAK Logistics, Inc. He leadsconsulting and software development projects to improve plan-ning and scheduling systems for clients in the pharmaceuticalindustry. Prior to joining SAK Logistics, Knight worked in theelectronics, metals, and automotive industries. Knight is a1992 Graduate Fellow from the MIT Leaders for Manufactur-ing Program. He holds a Master’s in Management Science fromthe MIT Sloan School of Management, a Master's in engineer-ing from the MIT Department of Mechanical Engineering, anda Bachelor’s degree in mechanical engineering from MIT. Hewas awarded the Goodyear Prize in Manufacturing Manage-ment from the MIT Department of Mechanical Engineering in1990.

SAK Logistics, Inc., 5335 Triangle Pkwy., Ste. 210, Norcross,GA 30092, (770)559-6396 or knight@saklogistics.com. Inquir-ies welcome.