Waste Identification Diagrams

CLME'2011 / IIICEM – 6º Congresso Luso-Moçambicano de Engenharia - 3º Congresso de Engenharia de Moçambique
Maputo, 29Ago - 2Set 2011 - Edições INEGI 2011, (ISBN: 978-972-8826-24-6), Ref: CLME’2011_0912A
J. Carlos Sá*1, J. Dinis-Carvalho2 and Rui M. Sousa2
Instituto Politécnico de Viana do Castelo (ESCE) – Valença, Portugal
Universidade do Minho, Departamento de Produção e Sistemas (DPS) – Guimarães, Portugal
*Email: [email protected]
ABSTRACT: The most popular and perhaps the most effective way to represent the material
flow in production units is the diagram known as Value Stream Map (VSM). Moreover these
maps are also used to help in the identification of waste as well as a tool to support
continuous improvement. Nevertheless, many of VSM limitations are known and thus there is
room for the creation of other more effective ways to represent productive units as well as
helping the identification of production waste. This paper presents a new graphic
representation model for production units, as a tool to identify three forms of waste,
designated as Waste Identification Diagram (WID), which aims to provide information to top
managers in a much more effective format. The WID is a network of blocks and arrows,
showing visually the throughput times, idle capacity, transport effort, changeover times and
work-in-process levels. To illustrate the main features of this new tool, the paper includes a
WID of a real production unity.
In 1988 Taiichi Ohno presented in his book “Toyota Production System: Beyond Large-Scale
Production" [Ohno, 1988] the seven wastes (muda) responsible for manufacturing systems’
low productivity. The types of waste were identified as: over-production, over-processing,
inventory, movements of people, transportation of materials, non-conformity of products and
people waiting. According to Taiichi Ohno, the existing wastes, along with inadequate flow of
production, are sources of losses and, therefore, need to be identified and eliminated. The
knowledge about the Gemba (shop-floor) is crucial to waste identification. Locations with
high inventory (probably the most “visible” waste) are not difficult to spot but, most likely,
that inventory is a consequence of other hidden wastes. Thus, the importance of
representation tools specifically designed to identify waste within a manufacturing system,
seems to be obvious.
The main purpose of this paper is the introduction of a new visual tool able to highlight three
types of waste associated with the items flowing on the shop floor, namely inventory,
overproduction and transport. This tool is designated as Waste Identification Diagram (WID),
and is developed by the Production and Systems Department at University of Minho,
The paper is organized as follows. After this introduction, section 2 provides a brief review of
some representation tools, for production systems, considered by the authors as relevant in the
context of this work. Section 3 describes the fundamentals of the developed tool (WID) and
introduces the main elements of the correspondent graphical notation. An application
example, based on a real industrial scenario, is presented in section 4. Finally, on section 5,
some concluding remarks are outlined.
CLME'2011 / IIICEM – 6º Congresso Luso-Moçambicano de Engenharia - 3º Congresso de Engenharia de Moçambique
Maputo, 29Ago - 2Set 2011 - Edições INEGI 2011, (ISBN: 978-972-8826-24-6), Ref: CLME’2011_0912A
Currently there are a number of methodologies to build graphic representations of production
systems, in order to help managers to describe, analyse and diagnose these systems. Value
Stream Mapping (VSM), introduced in ref. [Rother and Shook [1999], is probably the most
popular example. The VSM has revolutionized the graphic representations of production
systems, and its main objective is the representation of production and information flows,
where some of the existing waste (muda) may become easily identifiable (e.g. inventory and
over-production), especially by lean experts. Ref. [Serrano et. al 2008] state that VSM is
undoubtedly an innovative tool presented in the context of lean production. However, other
authors pointed out a number of limitations. For example, ref. [Gahagan, 2010] acknowledges
that VSM is a powerful tool to help the implementation of lean manufacturing, but refers, as
limitation, the difficulty in transmitting the results to other players when they are not familiar
with that methodology. According to ref. [Nazareno et al, 2003; Chitturi et al., 2007], the
VSM has major limitations in the representation of production systems characterized by high
diversity of products and several production routes. Finally, VSM is focused on materials and
information flows and does not represent the flow of people.
The flow process chart constitutes another type of graphical representation commonly used in
the area of industrial engineering [Courtois et al, 1997]. Flow process charts are intended to
represent the sequence of operations involved in the production of a given product, or family
of products. Specific symbols were created to represent processing operations, transportation
of materials, storage, buffers and inspection. Although not designed for that purpose, flow
process charts can identify some wastes, namely inventory and transportation of materials.
However, these charts do not include any kind of quantification, being thus of limited use in
the context of waste identification. Additionally, flow process charts are focused on the flow
of materials and the workers’ role is not addressed.
Somehow contrarily, the approach named “treasure maps”, described in ref. [Kobayashi,
1995] is focused on the workers’ role and does not address the materials flow perspective.
This approach creates a map with the location of wastes associated to workers’ activities
(mainly, movements and waiting) and it is based on work sampling studies conducted by
teams throughout the entire production system. “Gold mines”, “silver mines” and “cooper
mines” are designations used in the “treasure map” to identify the areas where waste occurs,
by decreasing order of degree of wastefulness [Kobayashi, 1995]. The main reason behind the
development of this approach was the fact that many workers, and managers, do not have a
clear understanding of what is waste.
Simulation is commonly applied in the design of production systems and to evaluate proposed
improvements to existing systems. This tool, due to its dynamic nature, allows the
identification of some wastes, namely, inventory, movements of people and transports of
materials. However simulation demands the utilization of specific software tools (most of
them proprietary) and a considerable level of expertise.
To overcome some of the limitations of the previously described approaches/tools, but mainly
to explore a different visual mechanism (innovative within the representation tools commonly
used in the production systems area), a new tool called Waste Identification Diagram (WID)
was developed in the Production and Systems Department at University of Minho, Portugal.
CLME'2011 / IIICEM – 6º Congresso Luso-Moçambicano de Engenharia - 3º Congresso de Engenharia de Moçambique
Maputo, 29Ago - 2Set 2011 - Edições INEGI 2011, (ISBN: 978-972-8826-24-6), Ref: CLME’2011_0912A
In this section it is presented a new powerful representation tool - WID - to help managers to
identify most forms of waste on their shop floors. WID allows the description of production
units, visually highlighting the main problems that prevent companies to achieve streamlined
production flows. The WID is easy to understand (due to the semantics of the developed
graphic notation), allows an immediate visual diagnose of the most relevant locations of waste
and can be used as a continuous improvement tool. With WID it is possible to represent not
only a single production route (of a given family of products), but many - the only limitation
is the size of the diagram.
3.1. Fundamentals
Two important fundamentals involved in the WID development process are the well-known
visual control and Little’s Law [Little, 1961]. The basic concepts of line balancing,
transportation effort and setup time were also considered.
Visual control is one of the tools associated to the lean production paradigm and intends to
provide an immediate visual perception about some relevant aspect (e.g. state of a production
process, work instructions, safety stock). Somehow, WID borrows this concept of immediate
visual perception and apply it to waste identification. In fact, as described in the next section,
the physical dimensions of the icons used to construct WID are proportional to some wastes
(e.g. inventory and transportation of materials). This constitutes the main distinctive
characteristic of WID when compared to other representation tools (section 2).
According to Little’s Law the throughput time of a given process can be obtained multiplying
the WIP (Work-In-Process) in that process by the Cycle Time (instead of Cycle Time the
authors use the Takt Time since it is more realistic in real production environments). WID
represent this multiplication (i.e. the throughput time) as the frontal area of a specific
tridimensional icon (Fig. 1) providing thus an immediate identification of the problematic
areas (the larger the area, the larger the throughput time).
The Waste Identification Diagram is basically a network of two types of icons: the block icon
representing workstations and the arrow icon representing transportation of parts.
3.2. Block icon
This is the main icon type on Waste Identification Diagrams (WID). Each block in the WID
represents one process or a group of processes (a workstation, a production cell or any other
production unit), connected to other blocks by transportation arrows. The size of each block
depends on four variables, namely, Takt Time (TT), Cycle Time (CT), Work In Process
(WIP) and Changeover time (C/O) (Fig. 1).
CLME'2011 / IIICEM – 6º Congresso Luso-Moçambicano de Engenharia - 3º Congresso de Engenharia de Moçambique
Maputo, 29Ago - 2Set 2011 - Edições INEGI 2011, (ISBN: 978-972-8826-24-6), Ref: CLME’2011_0912A
Fig 1. Block icon
The width of the block (X axis) is related to the WIP on that process. The units used to
measure WIP may be number of parts, weight units, length units, volume units or their value
(currency). The height of the block (Y axis) corresponds to the Takt Time and the height of
the green part is the Cycle Time of that process. The difference between the Takt Time and
the Cycle Time (CT), shown in orange, represents the unused capacity for that process. The
units used for TT and CT are time units (e.g. seconds, minutes or hours) per part.
The depth of the block (Z axis) represents the changeover time or setup time for that process
or workstation. When a process needs large setup times a natural and classical consequence is
large levels of WIP associated with it. In this way, may be expectable that thicker blocks
(blocks with high values of C/O) would also be wider (blocks with high values of WIP).
C/O=50 min
TT=90 s
TC=60 s
TC=30 s
TT=90 s
Idle capacity
C/O=20 min
One of the first interesting results of such representation is that, according to Little´s Law
[Little, 1961], the frontal area of the block (green area plus orange area) represents the
throughput time of the process. Thus, when observing the blocks, large areas represent large
throughput times.
Fig 2. Two block icons representing two different processes
In order to clarify the meaning of the visual information available on block icons, observe
Fig. 2. Assume that those blocks represent two different processes required in the process
route of a product or family of products (please note that the Takt Time is the same in both
processes since the Takt Time is product related). The first obvious information that can be
extracted simply by looking at the blocks is that the bigger is the block the higher is the waste
associated with it. Assuming that big blocks means big waste, Fig. 2 immediately reveals that
the WS_1 is creating a lot more waste than WS_2. Other obvious information coming from a
simple look at the blocks is that idle capacity exists more on WS_1 than on WS_2.
CLME'2011 / IIICEM – 6º Congresso Luso-Moçambicano de Engenharia - 3º Congresso de Engenharia de Moçambique
Maputo, 29Ago - 2Set 2011 - Edições INEGI 2011, (ISBN: 978-972-8826-24-6), Ref: CLME’2011_0912A
Other visual information can be easily obtained, such as:
Work Station 1 (WS_1) holds more WIP than Work Station 2 (WS_2).
WS_1 has more idle capacity than WS_2.
WS_1 and WS_2 are not balanced.
The lead time of WS_1 is higher than the lead time of WS_2. This is observed by the
frontal area of each block.
The Changeover time for WS_2 is higher than the changeover time for WS_1.
Based on the information available, the amount of WIP associated to WS_1 does not
make sense since both the changeover time and cycle time are low.
Broadly it may be stated that there is more waste associated to WS_1 even without a known
reason. It does not make sense so much WIP associated to WS_1 as its changeover time is
lower than the changeover time of WS_2 and also because WS_1 has plenty of spare capacity.
Just with this information, the manager would need to find out why the inventory associated
with WS_1 is so high. Sometimes there are important reasons but in other cases it happens
because no one is paying attention to it.
Another important piece of information coming out from these blocks is the throughput time
(lead time) of each process, i.e., the area of the front face of each block. As previously
mentioned the lead time, according to Little’s Law, is obtained multiplying the Takt Time by
the existing WIP. As can be seen from Fig. 2, the lead time associated to the WS_1 is 3 times
larger than the lead time of WS_2. Since the throughput time (and WIP) of WS_1 is so high,
special attention must be paid to that workstation in order to find out the reasons behind so
much inventory.
3.3 Transportation Arrow
Besides the block icon representing the two worse classic types of waste, which are inventory
and over-processing, being associated to a process, the WID also allows the representation of
another classic type of waste - the transportation waste, associated to the transportation of
parts between processes. The transportation arrow (Fig. 3) intends to represent the effort
needed to move the parts from process to process. The length of the arrow was decided to stay
constant for practical reasons since different arrow lengths could make the diagram difficult to
understand because of the scale problem. The thickness of the transportation arrow varies
according to the transportation effort associated to the connection from the supplier process to
the client process. The effort may be measured in parts*meter, kg*m, cost units (€) or any
other way of measuring transportation effort.
Fig 3. Transportation arrows with different thickness and measuring units
Since the thickness of each transportation arrow on the diagram is proportional to the
transportation effort, the thicker arrow on the diagram is associated to the transportation
segment with more room for improvement. In continuous improvement logic, actions should
CLME'2011 / IIICEM – 6º Congresso Luso-Moçambicano de Engenharia - 3º Congresso de Engenharia de Moçambique
Maputo, 29Ago - 2Set 2011 - Edições INEGI 2011, (ISBN: 978-972-8826-24-6), Ref: CLME’2011_0912A
be drawn for thicker arrows in order to make them thinner (leaner) probably changing layouts
approaching the corresponding client and supplier processes.
This section presents the application of the WID concept on a real production unit, more
specifically in a production section of a company dedicated to the manufacture of picture
frames. The information about cycle times, takt times, changeover times, and production
routes was given by the production department personnel while the information about work in
process (WIP) and travelling distances had to be collected by the authors on the shop floor.
The WID generated is presented in Fig. 4. The values of WIP in the diagram are the values
observed on a specific date and assumed as typical values representing normal condition.
120 Car*m
80 Car*m
80 Car*m
WIP = 4.680 m
20 Car*m
CT=3s TT=4,8s
WIP = 3.470 m
180 Car*m
WIP = 200 m
WIP = 9.580 m
WIP = 3.470 m
Fig 4. Example of a WID of a real production unit
Knowing that the icons’ size is correlated to the level of the associated waste it may be stated,
after a simple look at the diagram, that the PAINTING process is clearly the most critical spot
in terms of inventory and throughput time. It can be said that there is a lot of waste associated
to the painting process. The second most critical process in terms of throughput time is the
FILM M/C, although with much less associated WIP. In terms of transport, it is also clear
from a simple look at the diagram that the most critical spot is the transport from the FILM
M/C process to the INSPECTION process, with an effort of 180 car*m per day. The
PACKING process has very little amounts of WIP and some of the practices applied there
CLME'2011 / IIICEM – 6º Congresso Luso-Moçambicano de Engenharia - 3º Congresso de Engenharia de Moçambique
Maputo, 29Ago - 2Set 2011 - Edições INEGI 2011, (ISBN: 978-972-8826-24-6), Ref: CLME’2011_0912A
could probably be replicated to other processes. Another class of information that comes clear
on this diagram is the idle capacity available in all processes. Since the orange areas represent
idle capacity we observe that most processes are clearly underused. These diagrams can also
be used for continuous improvement, projecting future states for the diagram as it is done with
VSM diagrams. In general terms it may be said that any action able to reduce the sizes of the
blocks will result in better production performance. Another type of improvements can be
obtained by undertaking actions that result in the reduction of the orange areas.
The concept of Waste Identification Diagram (WID) was presented in this paper along with
its main features and advantages. These diagrams are able to describe production units with
precious visual information enabling managers in the identification of the main forms of
waste associated to the materials flow (inventory, overproduction and transport). The size of
the areas and volumes are proportional to the production waste so it is easy to identify the
locations of the most critical processes, i.e., with highest responsibility in generation of waste.
Indicators such as throughput time, idle capacity, transport effort, WIP and changeover times
are easily understood through a simple glance at the diagram. An example of a WID for a real
production unit was also presented highlighting the main locations needing improvements.
Chitturi R.M., Glew D.J., Paulls A., Value stream mapping in a jobshop. In: IET
(International Conference on Agile Manufacturing), ICAM 2007, Durham, 2007, 142-147.
Courtois A., Martin C., Pillet M., Gestão da Produção, 4ª Edição, Lidel – Edições Técnicas,
Gahagan S.M., Adding value to value stream mapping: a simulation model template for VSM.
Available at: http://www.iienet2.org/PrinterFriendly.aspx?id=7584 [Accessed 4 September
Kobayashi I., 20 Keys to workplace improvement. Rev. ed. Portland: Productivity Press, cop.
Little J., A proof for the queuing formula: L = λW, Operations Research, Vol. 9, 1961,
Nazareno R.R., Silva A.L., Rentes A.F., Mapeamento do fluxo de valor para produtos com
ampla gama de peças, XXIII ENEGEP – Ouro Preto, MG, Brasil, 2003.
Ohno T., O Sistema Toyota de Produção: Além da produção em larga escala. Bookman,
Rother M., Shoook J., Learning to See – value stream mapping to add value and eliminate
muda. The Lean Enterprise Institute, 1999.
Serrano I., Ochoa C., Castro R., Evaluation of value stream mapping in manufacturing system
redesign. International Journal of Production Research, 46(16), 2008, 4409-4430.