You are on page 1of 6

Available online at www.sciencedirect.

com

ScienceDirect
Procedia CIRP 19 (2014) 63 – 68

Robust Manufacturing Conference (RoMaC 2014)

Reliable Shop Floor Bottleneck Detection for Flow Lines through Process
and Inventory Observations
C. Rosera*, K. Lorentzenb, J. Deusec
a
Karlsruhe University of Applied Sciences, Moltkestr. 30, 76133 Karlsruhe, Germany
b
Robert Bosch GmbH, Postfach 30 02 20, 70422 Stuttgart, Germany
c
TU Dortmund Technical University, Leonhard-Euler-Str. 5, 44227 Dortmund, Germany
*
Corresponding author. Tel.: +49-721-925-1936; fax: +49-721-925-1947. E-mail address: christoph.roser@hs-karlsruhe.de

Abstract

Bottleneck detection in manufacturing is the key to improving production efficiency and stability in order to improve capacity. Yet common
bottleneck detection methods in industry and academia lack either accuracy or practicability, or both, for dynamic systems. The new
methodology is conducted by the observation of processes and inventories. Blocked processes and full inventories indicate a downstream
bottleneck. Starved processes and empty inventories indicate an upstream bottleneck. Through subsequent observations of multiple process
states and inventory levels within a system, it is possible to determine the direction of the bottleneck at the given time and hence to find the
momentary bottleneck in the system. The shifting of bottlenecks can be observed directly. Work-sampling techniques can be used to obtain a
long-term picture of the dynamically shifting bottleneck. The new methodology does not require any calculations, statistics, or time
measurements. Hence the method is suited for practical use by shop floor supervisors and clerks. The direct observation of the bottleneck also
gives additional information about the underlying causes of the bottlenecks, simplifying the improvement of the system capacity. Extensive
field testing of the method received positive feedback not only from management but also shop floor operators. The method is already in use at
the Robert Bosch GmbH, where it is known as the bottleneck walk.

©
© 2014 Published
2014 The Authors.byPublished
ElsevierbyB.V. ThisB.V.
Elsevier is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/3.0/).
Selection and peer-review under responsibility of the International Scientific Committee of “RoMaC 2014” in the person of the Conference
Selection
Chair Prof.and peer-review
Dr.-Ing. under responsibility of the International Scientific Committee of “RoMaC 2014” in the person of the Conference
Katja Windt.
Chair Prof. Dr.-Ing. Katja Windt.
Keywords: Bottleneck Detection; Shifting Bottleneck; Theory of Constraints, Variability

1. Introduction and Scope data-monitoring systems appropriate for bottleneck detection
for three reasons:
Bottleneck detection in manufacturing is the first and most
essential step to improve overall manufacturing capacity. Yet x Flow lines are often combinations of manual and automatic
as detailed in the paper below, existing methods lack either processes. However, live data of manual processes is
accuracy or practicability, or both. This paper aims to detect usually difficult to obtain and hence not available, even for
the bottleneck in flow lines. The presented methodology was the rare circumstances where this would be permitted by
developed by Roser at the Robert Bosch GmbH, where it is work councils.
known as the bottleneck walk. The method allows the x Not every station is equipped with a suitable electronic
continuous improvement of the system capacity. It is assumed system or an overall system network.
that the flow lines have defined buffers between processes and x Even if stations are equipped with data-monitoring
are not equipped with electronic data-monitoring systems. The equipment, the information gathered is usually insufficient
latter assumption is based on the authors’ practical experience, for bottleneck detection and lacks key information.
where most production lines are not equipped with electronic

2212-8271 © 2014 Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/3.0/).
Selection and peer-review under responsibility of the International Scientific Committee of “RoMaC 2014” in the person of the Conference
Chair Prof. Dr.-Ing. Katja Windt.
doi:10.1016/j.procir.2014.05.020

of the flow line. 2. Each process may at different times be blocked or starved. A accepted as a basic definition of bottlenecks for manufacturing simulation enables the user to model a system. Definition 3 is limited only to static systems. However. system and its dynamic processes in a software model that can Definition 5 is the one with the highest accuracy as proven by be experimented with in order to gain knowledge. Variations of this is the smallest among all the machines in the system. [7] finally defines the bottleneck as the process whose Approaches using utilization [10. alternatively. even if it has systems. But the 2. 3. / Procedia CIRP 19 (2014) 63 – 68 Therefore. 18] or related OEE sensitivity of the system’s performance index to its (Overall Equipment Effectiveness) measures enhance the production rate in isolation is the largest. In a production line. Bottlenecks Fundamentals In the next section. before searching for the bottleneck. the user is able to test the on the shop floor. Roser et al. The major accomplishment by [7] is the system under a variety of conditions [9]. method are for example the X-Factor theory [19] 4. Nevertheless these influences are in practice often negligible. Definitions 1 and 2 deliver a basic understanding of Time. For practical purposes please also note that a bottleneck does not necessarily have to be in a production process itself. systems. statistical data on process time are usually rather difficult to or neither blocked nor starved. Furthermore. improvement and bottleneck detection focus lies on the analysis of the gap between net production time and total time. even be a process within the information flow (regardless of push or pull systems). According to [7]. For practical bottleneck detection in the environment x Starved – a process has to stop because its preceding buffer described in this work. This method doesn’t include any losses and 3. making assumptions is one of the key or process is empty. while at the same time being general enough to be knowledge can then be transferred back to reality [8]. can be compared. Process Time first clearly define what a bottleneck is. [7] also observes that. Hence. but also logistics process that supplies processes. While average process times are often reasonably well known.64 C. albeit the author has also seen Simulation is an experimental procedure for modeling a instances where this influence could not be ignored. Utilization or OEE based Approaches largest work-in-process inventory in the preceding buffer. Therefore. The main flaw of this method is that it is based on bottlenecks. 4]. if the upstream process simulation can be excluded as a basis for a detection has a higher frequency of blockage than the downstream methodology. The all other processes. Afterward. recognized and described by several authors [1. This method 1. the described method not only contains the It can also be (and in our experience frequently is) in a method for evaluation of shop floor bottleneck data. the prerequisite for improving the bottleneck is to 3. offers a simple and fast way to detect the bottleneck. problems for the application of simulation software. then the process between the upstream and downstream processes is the bottleneck. . whereas definition 4 is only an indirect measure via inventory and hence subject to other influences resulting in flawed 3. Furthermore. A number of bottleneck definitions are available in literature: The process-time approach measures the process times in the material flow under isolated conditions. as compared to process-time approach by including performance losses. This [7]. 5. Hence the data quality is often insufficient for the level frequencies of blockage and starvation of adjacent processes of precision required for bottleneck detection. the obtain. buffer or process is full. Common Industry Bottleneck-Detection Methods find the bottleneck in the first place (bottleneck detection). 2. Simulation bottleneck detection. it is important to 3. the ability of the simulation software to visualize material flow design increases x Blocked – a process has to stop because its subsequent the system’s acceptance within the management [9].1. but rather the defined as the machine whose production rate in isolation maximum capacity under ideal conditions. [7] observes that on the shop floor a bottleneck is often therefore does not detect the bottleneck. proof that an evaluation of the processes being “blocked” or A simulation basically allows for detection of bottlenecks “starved” will find the bottleneck according to definition 5. But [7] also states that it cannot directly be applied not been built yet. common methods for bottleneck detection are discussed in the environment described in The importance of improving bottlenecks has been Section 1. but are not precise enough for shop floor averages and cannot detect shifting bottlenecks in dynamic application. process has of being starved. it can describes a process on how to raise the data on the shop floor.3.2. [6] calls it a resource whose capacity is lower than the method detects only the static bottleneck – the capacity limit demand or the process that limits throughput. [5] describes a bottleneck as a function that limits output. on the shop floor a bottleneck is often defined as the machine with the 3. especially when the combination of elements prohibits other These states can be defined as followed: classic bottleneck detection.

it may or may not be the bottleneck is probably upstream. This is obsolete if the machine state is obvious and will not change 3. it can be clearly stated when it is not the buffer is between one third and two thirds full. station happens. however. These methods work well and are able come from. For practical purposes. As with the processes above. but the chances of observing the direction become less likely. although in our experience they are infrequently is probably downstream where the parts go to. time. I. The average active period method x Blocked – Bottleneck is downstream. for parts or transport. However. while the active period method defines the process is waiting for parts (starved). it can be determined in which direction the bottleneck needs to be searched next. Of course other trade-offs are also possible. If the worked well. the 4. etc. then the process has an ongoing breakdown. the bottleneck literature. Scheduled break.4). then the bottleneck influencing that process must be capacity. If the From the authors’ practical experience. the bottleneck is overview of methods. and others [21] Please note. reliability. Observation of Inventories Overall. On the downside. It is important to acknowledge that for inventory evaluated in a systematic process (4. Set- period method [12] by the primary author are based on the Up. The result of these two levels around half capacity the bottleneck direction is highly steps is a ranking of bottleneck sets that limit the output of the uncertain. 4. observation alone if the process is the bottleneck. if used in industry. Other methods also look not only at probably upstream where the parts come from. Active Period Method overview of different possible system states and the conclusion about the bottleneck*. A clearly defined buffer can be filled between 0% and The Bottleneck walk is based on observations of different 100%. If a process is waiting *A first version of states can be found at [15]. The Bottleneck Detection Methodology or downstream † . If the operator is absent. Summary within the length of a process time. Of course also give hints to the direction of the bottleneck. it cannot be not enough information to assume a bottleneck upstream or the bottleneck.2) and inventory states (4. it is not important if the buffer is large If a process is waiting for transport (blocked).5). See for example [13.1. but rather if it is getting larger or smaller. While detecting the process state. The summary above focused on methods used in industry. then the or small. it may or may not be the bottleneck.. Furthermore. These environments. Maintenance. Roser et al.e. these methods require extensive downstream where these parts go to. process-related data that may or may not always be available. it is not absolutely certain that the momentary bottleneck is upstream 4. a one-third approach process is working. then the bottleneck is probably downstream. † for parts (starved). The process could work more but is slowed down by the bottleneck. If the buffer there are numerous other methods described in academic between two processes is full or rather full. tells the observer what the actual state is. C. and are hence unsuitable for industry. then the bottleneck must be upstream. downstream. that while this information is probable. then the bottleneck momentary bottleneck as the process with the momentarily influencing that process must be upstream where these parts longest active period. it may or may not be the thirds full. then there is bottleneck. it is only useful if the data is available. the bottleneck may be in either direction. the information is still relevant. duration a process is working without interruptions by waiting x Starved – Bottleneck is upstream. 20] for a recent the buffer is empty or rather empty. This data is gathered bottleneck is considered to be upstream. ‡ This is a further development of the 50% rule in [14] and [16].2. this is difficult bottleneck must be downstream. If the bottleneck.3. The closer the capacity is to one extreme the more likely 4. The average active period method [11] and the active x May be the bottleneck: Working. If the active period. Whenever the process is waiting. Observation of Process States the bottleneck is in the corresponding direction. . Hence around half capacity no valid statement can flow line during the period observed. The presented process time is essential to ensure precision. Breakdown. but also other objectives including throughput inventory is half full. Here it is necessary to decide at which point the process (4.5. WIP. from this observation of a waiting process. be made. defines the bottleneck as the process with the longest average The observation is taken directly at the process.4. Basic Methodology inventories can give us the direction of the bottleneck. waiting for the end of the As such. However. / Procedia CIRP 19 (2014) 63 – 68 65 3. If the buffer is below one third full. Hence a When observing a process. The collected data is unknown. If the process is waiting for transport of parts to determine the overall effect of processes on system (blocked). you would have to take the first derivative of inventories. process. Similarly. it cannot be determined by one tradeoff has to be made between accuracy and observability. applicable methods lack the ability to detect the shifting bottleneck for dynamic and instable shop floor The second source of information is the inventories.3). The list below gives an to observe reliably. and in practice the assumption above works well in our experience. The moment after bottleneck walk is based on these reliable methods while the process time ends and the transfer of the part to the next avoiding the extensive data requirement. or during a walk along the flow line (4. If the throughput. To reliably find the current momentary bottleneck. since the process is waiting on another downstream ‡ . If the buffer is above two bottleneck.

a transport process or another must be between the arrows pointing toward each other. it is also possible that 4 3Ź B Ż0 Ź 1Ź P Ż0 the bottleneck is outside the scope of the observations and the entire system may be slowed down by a lack of demand or 5 4Ź B 1Ź B 1Ź Ż Ż0 supply. for example. the calculation of the described above. the arrows line. in measurement 3 of Fehler! Verweisquelle konnte nicht 4.4. 2 2 Ź 1Ź Ź 1Ź P Ż0 Finally. above.6. and so on. and in about 1/3rd of the cases . in our experience even for systems with small observation and to try to understand why the process became buffer inventories of less than 5 pieces and rapid cycle times the bottleneck right then. secondary process that is the bottleneck at that time. For practical purposes. with a separate Verweisquelle konnte nicht gefunden werden. line 4 of Fehler! Verweisquelle konnte nicht gefunden the process states are abbreviated with “W” for waiting. Hence it appears that the process time of process 3 causes process 3 to become The Bottleneck walk passes along the observed flow line the bottleneck [14]. and monitors the data of different processes and inventories as For quantitative evaluation. bottleneck indicated in one line as. Hence the likelihood of a bottleneck shifting As above. The Walking Process processing a part when it is the bottleneck.5. This. Verweisquelle konnte nicht gefunden werden. Starved and Bottleneck 3 ŻŻŹŹ direction 1 ŻŹ 4. It may be that the bottleneck shifted just while you for processing. above. the observer walks along the the arrow tips may be the bottleneck. This simply means that the bottleneck is currently shifting. as shown in measurement 2 of Fehler! are listed in sequence on the top of the sheet. Roser et al. process 3 seems to be the most frequent bottleneck and is usually 4. Furthermore. The for the Bottleneck walk. All observed processes and buffers Secondly. bottleneck and should be focused upon in future improvement Of course with shifting bottlenecks it is possible that the activities. P Processing Ź Process Waiting. half of the cases the true bottleneck differed from the In addition. in practice a shift can also be points to remember. Blocked Buffer Capacity Max 0 1 2 3 4 4 ŻŻ ? ŹŹ B Breakdown Ż Process Waiting. However. it is also advised that the gathered sequentially (by walking) and not concurrently. First. there may be more than one observed during the walk. a bottleneck shifts happens less than later improvements of the bottleneck. while in theory this all sounds very while the processes involved are under observation is possible. Since this Circling the bottleneck with a red box visualizes the finding. and yet it is unknown which levels will yield consistent information about the bottleneck bottleneck process will eventually dominate the other direction. it is sometimes bottleneck frequency for each process is suggested. “B” for breakdown. “P” werden. sheet in the direction of the bottleneck. In roughly easy to determine where the bottleneck most frequently was. In the authors’ experience. Two or more processes are a Observing the waiting times of processes and the inventory bottleneck for a part of the line. all the processes between During the Bottleneck walk. The bottleneck then This may be. writing down the inventory levels and process states in may point to the gap between two observations as shown in one line of the data sheet each round. Similarly.. the column for every observed spot. shift of the bottleneck overlaps with the walk. for example. The Evaluation Process gefunden werden. the data sheet in Fehler! A B C Verweisquelle konnte nicht gefunden werden. happens rather frequently in practice. is not a fixed requirement evaluation divided by the total number of measurements. These insights will be invaluable for of less than 3 seconds. in for every buffer or process where the direction of the the authors’ experience. in practice it is helpful process with the largest bottleneck frequency is the primary to select the spots to be observed beforehand. as the data is To gain further information. Subsequently. the observations will also give clues to why a expectation of the management. and it will be different production lines to detect the bottleneck. is used. as in measurement 5 of Fehler! Verweisquelle 3 4Ź B Ż0 P Ż0 Ż Ż0 konnte nicht gefunden werden. In this case.66 C.. Examples Figure 1: Example data sheet for bottleneck detection Repeating a string of observations multiple times will give The authors have used this method successfully in over 20 a picture of the shifting bottleneck over time. It is the better to walk against the flow of material to avoid walking number of measurements the process was bottleneck by arrow “with” a single part. it is much more likely that there is a bottleneck can be determined. straightforward. In the example of Fehler! cannot be followed due to rounding problems. however. it will be one of the processes data sheet as shown in Fehler! Verweisquelle konnte nicht indicated as bottlenecks. gefunden werden. To combine these information bits into a picture. However. more than one process. Furthermore. The example shows a area between the arrows pointing to a bottleneck may cover common flow line with buffers in between. were walking past these two points taking data. a bottleneck process. an arrow is drawn on the data small process in between that has not been studied in detail. once per minute. observers look at the bottleneck immediately after each However. in reality there are again some additional but unlikely. above. / Procedia CIRP 19 (2014) 63 – 68 Especially in the case of small buffers the rule of 1/3rd often process became the bottleneck. above has a Max 4 Max 1 Max 1 Max 3 double vertical line between observation spots to remind the Index Buffer A Buffer B Buffer C Buffer D user about the possibility that there may be something else that 1 4Ź P 1Ź Ź 1Ź P Ż0 was not looked at in detail.

Application table in the lower part of Figure 2 shows the direction of the bottleneck and the bottlenecks in black for ten Bottleneck The method is based on multiple observations. impossible using the traditional methods like line-balancing The first example in Figure 2 shows a very basic case of an charts or average cycle times or inventories. process downstream. it was quite possible to observe short waiting time for material after almost every part. but through multiple ŻŹBottleneck Direction Bottleneck observations it was also possible to determine the likelihood Figure 4: Exemplary bottleneck-walk result of automotive component assembly line shown in Figure 3. workstations. Only the Bottleneck walk was was released from the stopper. the maximum buffer capacities of selected secondary transport-related processes and. Process C was frequently logistics process. In our experience. preparations. When the light turned green able to determine the bottleneck reliably in a minor process and there was no part at the stopper. with the parts transported via work-piece carriers and coupled by conveyor belts as shown in Figure 3. these stations had a very Despite the fast cycle time. minutes. and the quantities for bottleneck ignored by all other bottleneck detection efforts. Ź Ź Ź Ź Ź Ź Ź Ź Ź Ź Ź Ż Ż However. Plant management 2 Ź B Ż P Ż Ż Ż P Ż 3 Ź Ź Ź Ź P Ż Ż believed these quality-control processes to be the bottleneck 4 Ź B Ż Ż Ż Ż Ż Ż Ż based on cycle time and a large queue of material waiting for 5 Ź Ź Ź P Ź Ź B Ż these stations. C. between 30% and 50% of all bottlenecks are in such Similarly. being around 0. Figure 3: Material flow of automotive component assembly line Inventory more than 2/3rd full Figure 2: Fast-changing valve assembly line. The 4. The line consisted of different individual led to rapidly changing bottlenecks. not only was it possible to observe such rapidly Ż shifting bottlenecks in action. upon which the next part completely escaped attention. 7 Ź P Ź Ź Ź P Ż Ż Ż Quality 8 Ź Ź P Ź P P Ż Control 9 Ź B Ż Ż Ż Ż Ż Ż Ż 10 Ź Ź Ź Ź P Ż P Ż Q1 A B C D E F P Processing Ź Process Waiting. Bottleneck was detected in an unobserved of each process being the bottleneck. Nevertheless. rest position. the actual Bottleneck walks took only three where the arrows point to the bottleneck in a previously minutes each. Q2 B Breakdown Inventory less than 1/3rd full Ż Process Waiting. For the sake of clarity the The analysis shows that while there was usually a long line number of parts in each buffer has been omitted. / Procedia CIRP 19 (2014) 63 – 68 67 the bottleneck was in a previously unobserved secondary walks. with between 10 and 30 processes in the lines. Cycle times ranged from 2 seconds to 15 walk the bottleneck was very clear. noted directly in the material flow graph. This work-piece carrier was the bottleneck. with two other processes. In the This rapidly changing bottleneck was very easy to observe following. Index A B C D The second to last station consisted of two parallel quality- 1 Ź P Ź Ź Ź P Ż P Ż control processes for capacity reasons. the station was waiting for waiting times could be observed easily. two examples are presented to illustrate the using the Bottleneck walk. For the sake of clarity the bottleneck direction has been the bottleneck (50%). The number of observations is a core issue for the application of method . then the process was waiting for another control stations. Quality changed quickly. then the process was that was not even part of the investigation. However. In preparation for the Bottleneck time was barely noticeable. despite significant effort to reduce 6 Ź P Ź Ź Ź Ź Ź B Ż the process time. before the quality control stations. Starved. The assembly line producing electronic components needed to be combination of fast and similar cycle times with small buffers improved. The cycle time was very fast. being occasionally the bottleneck (30% and 20% respectively). often only three to five parts. are buffers were measured. we selected one or two spots at each station where the second cycle time. material despite the long queue of material before these For example. It turned out that a small device was moving the work- Whenever there was a small delay between the arm returning piece carriers to one or the other of these two parallel quality- and the part moving. waiting for material upstream. produced every two seconds. Nevertheless. the part started moving to the next station. with the bottleneck moving to a different Control process roughly every ten minutes. with one part being observation. the bottlenecks unobserved spot between the buffer and the quality control. The last process was never the bottleneck. the overall capacity did not improve. as such. These methods assembly line for a valve. After these an exemplary Bottleneck walk is shown in Figure 4 below. but would have been difficult or procedure and its advantages compared to other approaches. but miss bottlenecks processes that are not under cycle times. Blocked. The buffers between the stations In the second example. This valve is assembled on a fully find bottlenecks only in the processes that are directly automated line with four major stations. the capacity of a highly automated were very small. during Ż almost every walk.2 seconds of a 3- walk. The waiting times in processes.7. Due to the nature of the system. it has so far small light going from red to green. for each logistic process. A and D. The result of upstream/downstream/unknown were decided. While not all data points gave a direction. Roser et al. Or a verification process ended with a As an otherwise insignificant secondary process. the actual bottleneck was very clear. each having similar observed. when an arm adding a spring returned to its parallel stations.

2010 accuracy.... Bernd... pp. M. In: Mathematical Problems in Engineering 2 (1996). 2008 Bottleneck detection is a critical part of the continuous [9] Nyhuis. 1998 of the averages is reduced and the observation is repeated [16] Lorentzen. 5 (2011): 391–404. K. 1990 observation are given from shop floor experience. The key advantage of the Bottleneck walk. the bottleneck walk Professional. .: Bottlenecks in serial production lines: A systemtheoretic approach. 443-455 with shorter system cycle times. Winter Simulation Conference. Bottleneck detection methods Aachen.68 C. systems through practiceoriented bottleneck management.. besides its Como. G. Conclusion [8] VDI: VDI-Guideline 3633 Page 1: Simulation von Logistik-.. lacking either validity or usability or both. bottleneck methodology on types of manufacturing other than For these managers. J. pp. 3. commonly used and New York: Springer. 2012 that use averages overlook shifting bottlenecks.: Simulation Modeling and Analysis. J. Edition. they often have the opportunity for a for a transfer of this methodology. 6. edpp. Further research groups: first. production lines – a focus on interdeparture time variance. Upper Saddle River: Prentice Hall. Shifting Bottleneck Detection.. / Procedia CIRP 19 (2014) 63 – 68 described. A. the 4th European Conference on Technology Management.-T. In E. and Huaxin Liu. [7] Kuo. Proceedings of Japan. In [11] Roser. Italy. P.. H. Tanaka. If the duration [15] Kuhn. For accurate Diego. yet this is a key requisite for bottleneck [14] Lorentzen. [3] Goldratt. In: Fördern und Heben (F+H) 51 (2001). Pei B: A dynamic system regulation measure thoroughly field-tested in different manufacturing plants. Tanaka. J. Wiesbaden: Gabler.: Computerized shop floor scheduling. 9. 2009 can also be easily taught to shop floor operators even without [18] Schweitzer P.: 2002. 336- that produces a part every 2 hours. Furthermore. USA: Institute of Electrical and of the line believed. N. C. Materialfluss und Produktionssystemen.: Produktionsmanagement. Glasgow. 949-953. two or three observations per day are a flow lines. M. TU Dortmund University. 2012 Detecting bottlenecks in serial momentary bottleneck before making statistical conclusions. Aquilano.. Since the process states can be easily obtained in viable approach. Advanced providing a reliable and practical way to find the bottleneck. E.. industrial engineers. Ritzman. In: International x The frequency of shifting bottlenecks increases for systems Journal of Production Research 26 (1987). Maschek. The bottleneck walk is effective adaptiveness and superior result.Lecture Notes in Computer improvement cycles as demanded by the concept of lean Science Volume 729. 50 No.: 2001. Berlin and Heidelberg improvement process. M. P. Nr. Performance Evaluation of knowledge of mathematics. Since those managers frequently cross the flow lines. Performance improvement in production the Toyota Central Research and Development Laboratories. Lim.: Das Ziel: Eine Methode ständiger Verbesserung - Höchstleistung in der Fertigung. T. and secondly. Germany. Further research will concentrate on the extension of the frontline managers (first level of leadership). detection in dynamic systems. Deuse.e. Hwang Y. Homewood: Irwin. and the Robert Bosch GmbH. 5. Kreutzfeldt J. pp. For this reason. The presented Bottleneck walk provides a framework for a [12] Roser. Electronics Engineers. K. 1079-1086. McGraw-Hill Science. 1991 practical use. edpp. for increasing effective capacity: the X-factor theory. single measurement using the Bottleneck walk methodology. The following practical rules for frequency of [2] Goldratt. This methodology allows gathering information in a structured References way. “Modelling Dynamic Bottlenecks in Production Networks. In: Proceedings of APMS2010 – International Conference on Advances in Production Management Systems. A Survey of Bottleneck Analysis in Closed Networks of Queues. Winter Simulation bottleneck was in a process different from what the managers Conference. pp.. 338 [5] Krajewski. Serazzi. M. is its simplicity. pp. Broglia M. Richter. K. J. J.. CA.” International Journal of Computer Integrated Manufacturing 24. New York: Mcgraw-Hill and can be quickly applied... I. a system that produces [4] Gutenschwager. Edition.. Berlin: Beuth. Dissertation. 2009 [21] Scholz-Reiter.: Simulation in der Lagerlogistik: Anforderungen und a part every 5 seconds shifts more frequently than a system Möglichkeiten. USA. C. Heidelberg: Physica. 3. no. pp 491-508 production [17] and discussed in [14]. L. Analysipp. Edition.. Semiconductor Manufacturing Conference and Workshop. London: McGraw-Hill. J. 6. 233-276 5. Nakano. J. Virginia. A Practical Bottleneck two thirds of all bottleneck detections done by the authors. 1993. Meerkov. Roser et al. then the effect of inventory buffers will likely for variability reduction as a roadmap for continuous improvement in diminish the accuracy of the observations. flow production. E.: Systematic approach frequently. Nr. M.: Production and operations management: A x Frequently shifting bottlenecks allow for more frequent life cycle approach. the method can have different target 6. M. For this. 4th European Conference on Technology Management. bottleneck detection it is necessary to determine the [13] Betterton C. Nr. whose dedication is to personally improve manufacturing systems. pp.: Operations management: Strategy and x The more balanced a the cycle times of the processes are. Therefore. C.: Systematische Variabilitätsreduktion zur kontinuierlichen Verbesserung von Fließlinien. Silver S. Cox. International There are few methods that can detect the momentary Journal of Production Research V. The method was [19] Delp D.. Katja Windt. R. it allows for a focused approach based on shop floor observations without requiring a large investment [1] Adam. 2001 of time. Si J.-T. 2008 bottleneck detection methods are woefully inadequate for [10] Law. 2002 the larger the frequency of bottleneck shifts.. 1. Edition. In B. E. necessary for this approach. it enables quick Computer and Communication Systems. Unfortunately.. 1992 bottleneck detections.: Toyota Kata: Managing people for improvement. the Detection Method. San simple yet accurate bottleneck detection method. L. Nakano. Peters et al. M. D.. 15 pp 4158–4174 bottleneck reliably. No stopwatches or formulas are [17] Rother. B. Yucesan et al. Shaker. R.: Fließproduktionssysteme. A. Arlington. other manufacturing types are the next challenge manufacturing site. M.. Cernobbio. [6] Chase. 2003 IEEEI/SEMI pg 81-88 The method has been developed based on research done at [20] Schultheiss J.: Beiträge zu einer Theorie der Logistik..