Chapter 2

General Disassembly Process

Abstract The economic feasibility of the disassembly process is a main issue

restricting its implementation in industry practice. Much research in the planning
of disassembly processes and operations has been conducted in order to increase
its economic feasibility. This chapter presents various aspects of the disassembly
process including product representation, disassembly sequence planning (DSP),
and dismantling techniques. This general knowledge is not limited to manual dis-
assembly, but is also useful in automatic disassembly, which is presented in the
following chapter.

2.1 Disassembly Process Planning (DPP)

The disassembly process is generally economically infeasible due to the diffi-

culties in the process. Designing products according to Design for Disassembly
(DfD) guidelines [1, 2] is expected to resolve this problem by making the disas-
sembly process easier. However, few products nowadays are actually designed
according to DfD. Therefore, the disassembly process remains difficult for the
majority of products. Hence, this book focuses on the means of improving the eco-
nomic feasibility of disassembly apart from DfD.
Duflou et al. [3] summarise the factors that influence profitability of the disas-
sembly process. Two major factors which are further explained in this book are
the (a) completeness of disassembly and (b) degree of autonomy of the process.
The desired completeness or depth of disassembly is a question addressed in disas-
sembly process planning and disassembly sequencing, and is further explained in
Sect. 2.2. The degree of autonomy can vary from complete manual disassembly to
semi-automatic disassembly and fully-automatic disassembly. Since automation is
a major theme of this book, an overview is explained in Chap. 3, with other details
presented throughout the rest of the book.
This chapter gives an overview and literature review regarding disassembly
process planning.

© Springer International Publishing Switzerland 2015 9

S. Vongbunyong and W.H. Chen, Disassembly Automation, Sustainable Production,
Life Cycle Engineering and Management, DOI 10.1007/978-3-319-15183-0_2
10 2  General Disassembly Process

2.1.1 Difficulties in Disassembly

The disassembly process cannot be considered as the reverse of assembly. This is

largely due to increased uncertainties: the disassembly process deals with unpre-
dictable characteristics in both the quality and quantity of EOL products. This
causes disassembly to be more difficult than assembly in the following aspects.
Uncertainties within models
Gungor and Gupta [4] summarise the physical uncertainties that can be found in
EOL products manufactured under the same model. These uncertainties result
from (a) component defects, (b) upgrading or downgrading during usage and (c)
damage during the disassembly operation.
Defective main or connective components can result in difficulties during the
removal operation which range from undesirable to dangerous. Examples include
chemical leakage from batteries and broken fasteners that cannot be disestablished
using common disassembly tools.
Upgrading and downgrading of the product during the usage stage can
result in a change in product and component configuration. This situation is
commonly found in devices containing exchangeable modules such as the per-
sonal computer (PC). Repairs and upgrades, e.g. involving the installation of
random access memory (RAM) or graphics card, are common during the usage
stage.
Damage in the disassembly process potentially occurs when the returned prod-
uct is fragile. The disassembly process may need additional steps or a change in
the disassembly sequence when certain parts are likely to break during the process.
Model-related variations
Products are manufactured into different models within the same product family.
Different models contain variations in characteristics, including material, size
and internal configuration. The same model may also be sold under differ-
ent brands. Optimally, model information should be obtained from a product
design database, taking the form of well-documented product specifications
or a Computer-Aided Design (CAD) model. Unfortunately, this informa-
tion is usually unavailable by the time EOL products are returned. Therefore,
the disassembly process needs to deal with incomplete product information,
some of which is only revealed during the disassembly process. The challenge
is to develop a disassembly plan that is general enough to deal with these
­uncertainties [5].
Difficulty of operations
Kroll et al. [6] define the term disassemblability to quantify the ease of product
disassembly. A product is assessed for disassemblability according to the dif-
ficulty of the disassembly operation, by assessing it against five major criteria:
(a) c­omponent accessibility, (b) precision in locating the component, (c) force
required to perform tasks, (d) additional time, and (e) special problems that cannot
2.1  Disassembly Process Planning (DPP) 11

be categorised in the other areas. Mok et al. [7] summarise the characteristics of an
ease of product disassembly as follows:
• Minimal force exertion;
• Quick operation without excessive manual labour;
• Simple mechanism of disassembly;
• Minimal use of tool: ideal disassembly should be performed without tools;
• Minimal part repetition: parts easy to identify at each state of disassembly;
• Easy recognition of fasteners;
• Simple product structure; and,
• Avoidance in usage of toxic material.
Gupta and McLean [8] state that the development of optimal disassembly plans
relies on four key phases: (a) product analysis, (b) assembly analysis, (c) usage
mode and effect analysis and (d) dismantling strategy. Firstly, the product must be
analysed and represented systematically. Options regarding the disassembly pro-
cess can be generated from or represented using the product structure. The process
can be considered at two levels, which are the sequence plan and the operation.
The completeness of disassembly is considered a part of the sequence plan.
In summary, the uncertainties and variations found within returned products
leads to uncertainties in the disassembly process. Uncertainties and variations of
the disassembly process are summarised in Table 2.1.

Table 2.1  Summary of variations and uncertainties in the disassembly process

Category Specific issues
Uncertainty in EOL condition • Modification of product during usage phase
• Condition of product
• Condition of main component
• Condition of connective component
Diversity of the supplied products • Main product structure
• Physical appearance of components
• Quantity of components
• Location of components
• Variation in manufacture (quality)
Complexity in process planning and • Disassembly sequence plan
operations • Disassembly operation plan (considering
prior actions)
• Disassembly process parameters
• Capabilities and limitations of detection
technique
• Precision of robot’s sensors—actuators
External factors • Technology and design changes
• Market driven factors
12 2  General Disassembly Process

2.1.2 Representation of Product Structure

The structure of a product consists of (a) components and (b) connections [9].
A component is an element that keeps its extrinsic properties, i.e. functionality and
material properties, after being detached from the product. A component cannot
be further dismantled without using destructive disassembly methods. The connec-
tion or liaison is a relation that physically connects two components to restrict the
motion between them. The task of disassembly is to disestablish these relations in
order to separate the relevant components.
Fasteners
A fastener is a component or design element that serves the purpose of connecting
other (main) components. Fasteners that are insignificant to the goal of disassem-
bly may be modelled separately to the main components. Lambert and Gupta [9]
­considers such fasteners as quasi-components, which can be discrete components
(e.g. screws, rivet, cable, etc.) or part of the main component (e.g. snap-fits).
Connection-establishing elements, such as solder and weld joints, that do not form
a component in themselves, can be considered virtual components.
Product structure
The structure of a product can be represented in many ways, of which two will be
detailed here: the connection diagram and the disassembly matrix.
First, the connection diagram (liaison diagram) graphically represents the com-
plete product structure using an undirected graph. The components are represented
by nodes and the connections by arcs. According to the level of detail, the graph
can be shown in three different forms: (a) extended form (b) reduced form and (c)
minimal form (see Fig. 2.1).
In Fig. 2.1a, the product is a composition of three main components—A, B, and C.
A and B are connected by mating. B and C are connected with a screw E which is
considered a quasi-component. C and A are connected by a weld D which is a vir-
tual component. The extended form shows full details of the product with every
­component and fastener. All fasteners, including virtual connections, are modelled
(see Fig. 2.1b). The reduced form represents the structure more concisely by hiding
the virtual components and using dashed lines for quasi-components. In this case,

Fig. 2.1  Connection diagram [9]. a Product assembly. b Extended form. c Reduced form.

d Minimal form
2.1  Disassembly Process Planning (DPP) 13

the connections associated with the virtual component, D–A and D–C, are removed.
As a result, only connection A–C is retained representing the weld (see Fig. 2.1c). The
minimal form shows the structure of the product in the most compact way by hiding
both virtual and quasi-components. This form represents the product in the simplest
way while preserving the information regarding the main components (see Fig. 2.1d).
Second, the product structure can be represented by a disassembly matrix, which
a computing approach (e.g. Linear Programming (LP) or Integer Programming (IP))
can be used to solve the disassembly planning problem. The disassembly matrix is an
N × N connectivity matrix where N is the number of the components. Each element
of the matrix represents the existence of a connection between two corresponding
components: “1” if a connection exists, and “0” if it does not. This information is
completely represented by the lower left part of the matrix, since the matrix is sym-
metric and the elements on the diagonal axis are non-applicable. From this matrix, it
is clear that the maximum number of connections is ½(N)(N−1). The disassembly
matrix of the example product in Fig. 2.1a is shown in Eq. (2.1).

A B C D E

B 1
Disassembly Matrix =
C 1 1

D 1 0 1

E 0 1 1 0 (2.1)

2.1.3 Disassembly Process Representation

The steps of the product disassembly process and their corresponding relationships
can be schematically represented in many ways. Lambert and Gupta (2005) [9]
summarise these approaches as follows:
Disassembly precedence graph
The disassembly precedence graph expresses sub-tasks of the disassembly process
connected and constrained by precedence relationships. This can be represented
in two forms: as a component-oriented or task-oriented graph (see Fig. 2.2). The
arrows communicate the ordering in which tasks must be performed. This tech-
nique was originally used for assembly process representation and assembly line-
balancing problems. Gungor and Gupta [10] introduce this to the disassembly
process due to its simplicity. However, a major disadvantage is that a complete
disassembly sequence cannot be expressed in one graph [11].
Disassembly tree
The disassembly tree expresses all possible choices for disassembly sequences,
and is derived from a table containing all possible sequences sorted by level and
14 2  General Disassembly Process

Fig. 2.2  Disassembly precedence [9]. a Assembly. b Component-oriented. c Task-oriented

Fig. 2.3  Example product Bourjault’s ballpoint [9]. a Assembly. b Connection diagram

operation type. A widely-used example is the Bourjault tree [12]. Two major
drawbacks are the complexity arising in complex products and difficulty in rep-
resenting parallel operations. Figure 2.4 shows a Bourjault tree representation of
the disassembly process of a sample product, the Bourjault’s ballpoint, which is
shown in Fig. 2.3. This product will also be used to demonstrate the representation
methods described in the following sections.
State diagram
The state diagram represents the disassembly sequence as an undirected graph,
where each node represents a state of disassembly. This can be categorised into
two approaches: (a) connection-oriented [13] and (b) component-oriented [14,
15] (see Fig. 2.5). All possible combinations of connections are represented by the
nodes. Each edge represents the establishment or disestablishment of a connection.
The major advantages are that the disassembly sequence of the complete product
can be demonstrated in one diagram, and the diagram is compact even for complex
products. However, state diagrams are unable to show how the disestablishment of
some connections cannot be done individually without affecting a combination of
related connections.
Kara et al. [16] used a connection-oriented state diagram representation to
develop a graphical representation method, the disassembly-sequence diagram, for
representing the disassembly sequence to and from different stages of the process
for selective disassembly. This diagram can be automatically generated from the
liaison and precedence relations. An example is shown in Fig. 2.6.
2.1  Disassembly Process Planning (DPP) 15

Fig. 2.4  Disassembly tree of the Bourjault’s ballpoint [9]

Fig. 2.5  State diagram of the Bourjault’s ballpoint [9]

AND/OR graph (Hypergraph)

This graph represents disassembly sequences based on subassemblies. A process
is represented by multiple-arcs (hyper-arcs) pointing from a parent to its child
components (subassemblies) (see Fig. 2.7). This overcomes the drawback of the
state diagram. However, a major drawback is the complexity of the visual repre-
sentation, which may become difficult to read when the number of components
increases. Lambert [17] proposes a simplified version of this graph named the
concise AND/OR graph. Further developments, aimed at representing the product
model and its constraints more accurately, include the arborescence with hyper-
graph [18], Petri net [19], and Hybrid graphs [20].
16 2  General Disassembly Process

Fig. 2.6  Disassembly-sequence diagram [16]. a Liaison diagram. b Diassembly-sequence

diagram

Fig. 2.7  AND/OR graph of
the Bourjault’s ballpoint [9]

2.1.4 Disassembly Sequence Planning (DSP)

A disassembly sequence is a procedure for the disestablishment of connections

and detachment of parts in the disassembly operation. The initial state is defined
as the complete product, and the final state, as a state where all desired compo-
nents or subassemblies have been separated. The main purpose of disassem-
bly sequence planning (DSP) is to find the optimal sequences of disassembling
products with respect to certain factors, e.g. cost-effectiveness, material return,
component recovery, and duration of operations. Theoretically, the number of pos-
sible sequences increases exponentially according to the number of components.
Therefore, finding the optimal solution is considered an NP-complete optimisation
problem [4].
Lambert [5] summarises effective methodologies based on a product-oriented
approach as follows. As adaptability is required for a flexible automatic disas-
sembly system, the main theme of this book, emphasis is placed on the adaptive
planners.
Mathematical programming (MP) method
The mathematical programming (MP) method aims to make the internal variables
converge to their optimum value without considering the complete search space.
The problem model is derived from a hypergraph (AND/OR graph). Costs are
assigned to each action (arc) with respect to subassembly components (i.e. parent
2.1  Disassembly Process Planning (DPP) 17

and child) and stored in a transition matrix. This can then be effectively solved
by mathematical solvers, e.g. using Linear Programming (LP), Mixed Integer
Programming (MIP), or Dynamic Linear Programming (DLP). Petri nets are also
used in case of a dynamic approach.

Heuristic methods
Gungor and Gupta [21] present a heuristic algorithm used to find near-optimal
solutions to the disassembly sequencing problem. Near-optimal solutions are
considered instead of optimal solutions, which are sometimes difficult to find
due to the size of the search space. This method requires information of the
precedence relationship among each of the components and the difficulty in
performing each action. Efficiency is evaluated by the authors based on disas-
sembly time. A case study regarding the DSP of a cell phone using the heuristic
method and different search algorithms, e.g. greedy k-best and A*, is examined
by Lambert and Gupta [22].

Artificial intelligence (AI) methods

Various techniques are used in artificial intelligence to generate and utilise con-
straints and reduce the size of the search space. Lambert [5] reviews typical AI
techniques for disassembly sequence planning, including simulated annealing
algorithms, genetic algorithms (GA), fuzzy sets, neural networks, multi-agent
­systems, and Bayesian networks. Other novel algorithms that have been efficiently
applied to DSP include ant-colony optimisation [23], case-based reasoning [24]
and rule-based sequence generation on clustering graphs [25].
Adaptive planner
An adaptive planner generates a disassembly sequence with respect to the uncer-
tainties and unexpected circumstances encountered during the disassembly
operation. Due to its particular relevance to automated disassembly, a number of
publications relating to adaptive planners have been reviewed in this section. The
literature handles the problem at two levels: the (a) process planning level and (b)
operation level.
In the process planning level, Tang [26] proposes using a Fuzzy Petri net to
model the dynamics of disassembly, including the uncertainties in product condi-
tion and human factors. The system is trained with data and feedback from the
actual disassembly, and selects the appropriate disassembly plan based on past
experience. Turowski et al. [27] presents an implementation of a Fuzzy Coloured
Petri Net for balancing a disassembly line. Grochowski and Tang [28] propose a
learning approach using a Disassembly Petri Net (DPN) and Hybrid Bayesian net-
work. Veerakamolmal and Gupta [29] propose using case-based reasoning (CBR)
to generate disassembly plans for multiple products. The plan for a new product
is adapted from existing plans by deriving it from a base case. Gao et al. [30] pro-
pose using a Fuzzy Reasoning Petri Net to adaptively generate the d­ isassembly
sequence according to the condition of the product observed at each state.
Decisions are made based on the estimated value returned, hazard level, and disas-
sembly cost.
18 2  General Disassembly Process

In the operation level, Salomonski and Zussman [31] propose using a predictive
model with DPN to adaptively generate the disassembly process plan according to
real-time measurements conducted by a robot arm. Lee and Bailey-Van Kuren [32]
address the uncertainties in the operation level by automatically recovering from a
visually-detected error. In addition, Martinez et al. [18] propose a dynamic sequence
generation method that generates an optimal disassembly plan during operations in
response to unpredictable situations, e.g. failure to remove a corroded part, replace-
ment of screws, etc. This system is modelled and controlled by a multi-agent system
(MAS). ElSayed et al. [33] use GA to generate an optimal disassembly sequence
according a supplied bill-of-materials (BOM) and components detected in real time.
Relations defined in the original BOM must be preserved.
In conclusion, the existing adaptive planners deal with many types of uncer-
tainty experienced during the disassembly process. The uncertainties relate to vari-
ations in the component conditions that deviate from the ideal case. The ability for
an adaptive planner to handle these uncertainties stems from its ability to appropri-
ately adapt existing knowledge into a new plan according to sensed information.
Machine learning techniques are used to allow the system to improve its perfor-
mance from past experience. However, the structure of the product, e.g. BOM and
CAD model, generally needs to be supplied a priori. A methodology accounting
for an uncertainty in the general product structure in real-time has not yet been
proposed in any research. In addition, the learning process has only been imple-
mented in the planning level. Hence, learning at the operation level, such as in
optimising process parameters, should be further investigated.

2.2 Completeness of Disassembly

The completeness of disassembly can be categorised into two types: (a) complete dis-
assembly and (b) incomplete disassembly. A complete or full disassembly is the pro-
cess that separates every single component of the product. This is rarely done due to
the technical constraints (particularly the complexity and the uncertainties in the opera-
tion) and high labour costs. On the other hand, the incomplete or selective disassem-
bly separates only the desired components or subassemblies, and terminates when the
desired depth of disassembly is reached. Disassembly becomes more cost-efficient with
a strategic choice of disassembly targets. Reasons for selective disassembly include
recovering modules or components for use as spare parts, separating those that contain
hazardous substances, and improving the quality and quantity of shredder residue [17].
Figure 2.8 illustrates the situation in maximising the profit in the disassembly pro-
cess. The disassembly range refers to the completeness of disassembly. The cost of
disassembly is due to operation time, varying according to the number and type of
connections to be disestablished. This increases with the desired completeness of dis-
assembly. Disassembly is economically feasible when the total profit from treating
or recycling all products exceeds the cost of disassembly. The optimal strategy is the
point at which the maximum profit can be obtained [34, 35].
2.2  Completeness of Disassembly 19

Fig. 2.8  Determination of optimal disassembly strategy [34]

The outcome of selective disassembly can be one of the three following types [9].
• Homogeneous components: parts that cannot be physically disassembled.
• Complex components: components comprised of a number of homogeneous
components, joined together with fasteners, which can only be separated using
destructive disassembly.
• Modules: sets of components that perform a self-contained function. Modules
can be further disassembled via non-destructive or semi-destructive operations.
However, maintaining their original condition and functionality can allow reuse
of the entire module.
The researchers currently focus on developing a methodology to find optimal disas-
sembly sequences in which the completeness of disassembly is taken into account.
Kara et al. [36] propose the methodology of developing the optimal selective disas-
sembly sequence which is the reverse of the methodology for assembly presented
by Nevins and Whitney [37]. The disassembly sequences are generated from the
product specifications, namely list of parts and subassembly, precedence rules,
product representation model, and disassembly sequence diagram. Subsequently,
the optimal sequences for removing the selected parts are obtained by removing
invalid sequences according to liaison analysis. In regard to this concept, software
that automatically generates and visualises optimal sequences of selective disassem-
bly from specified constraints is developed by Kara et al. [16, 38].

2.3 Disassembly Operations

2.3.1 Types of Fasteners

The disassembly operation is divided into two main tasks: disestablishing fasteners
and detaching main components. The main component is detachable if the associated
20 2  General Disassembly Process

Table 2.2  Disassembly methods according to fastener type

Discrete components Not deformed Bundling Shear cut
Spring Deform/pull
Screw, bolt, nut, washer Unscrew, drill
Reversibly deformed Cotter pin, staple Pull
Irreversibly deformed Rivet Pry out, drill
Adhesive: glue, seal Peel, pry out, break
Parts of components Reversible connection Surface: mating Remove
(semi-reversible) Surface: press fit Pull, pry out
Snap fit Deform, pry out,
pull
Irreversible connection Surface: press fit Pull, pry out
Surface: mould Break
Seam fold Deform
Seal Peel, pry out, break
Virtual components Irreversible Solder Shear cut, break,
melt
Weld Saw cut, break

connections are located and disestablished. Specific techniques are required for
effectively disestablishing different types of fasteners. Lambert and Gupta (2005)
[9] categorise the fasteners commonly found in mechanical and electronic-electrical
products into 13 types. Different types of fasteners require different disestablishment
methods and display different levels of reversibility. A summary of fastener types
and their respective disassembly methods is shown in Table 2.2.

2.3.2 Dismantling Techniques

Disassembly operations can be broadly categorised into three types: (a) non-
destructive, (b) semi-destructive, and (c) destructive disassembly. The characteris-
tics of each category are explained in detail as follows.
Non-destructive disassembly
All outputs of non-destructive disassembly remain undamaged. This is desired for
maintenance, component reuse and remanufacture. All fasteners within the p­ roduct
must be reversible or semi-reversible. The dismantling of reversible fasteners (e.g.
screws) is generally easier than that of semi-reversible fasteners (e.g. snap-fits).
The operation cost is generally high, as high flexibility is required, particularly due
to difficulties such as rust and partial damage. Even though a number of tools have
been specially developed to facilitate non-destructive disassembly, e.g. for the dis-
assembly of screws [39] and snap-fits [40], the non-destructive approach is still
generally economically infeasible [3].
2.3  Disassembly Operations 21

Semi-destructive disassembly
The semi-destructive approach aims to destroy only connective components, e.g.
via breaking, folding or cutting, leaving main components with little or no dam-
age. This increases the efficiency of the operation and has been found in many
cases to be economically feasible. Many research works relating to automatic
disassembly use semi-destructive techniques to overcome the uncertainties in the
product condition and geometry. Examples of such techniques include the drilling
out of screw heads during the disassembly of electric motors [41], creation of new
surfaces allowing torque transmission for unscrewing [34] and cutting off of screw
heads using a cut-off wheel [42].
Destructive disassembly
Destructive disassembly deals with the partial or complete destruction of obstructing
components. Components or irreversible fasteners, e.g. welds, are destroyed using
destructive tools such as a hammer, crowbar or grinder. These operations are fast,
efficient and inherently flexible. As a result, destructive disassembly is economically
feasible and commonly performed in industry practice. One common application of
destructive disassembly is in the opening of a covering component to reach the more
valuable components inside. Examples include the breaking of the separating line
[34] and using plasma arc cutting to destroy the metal casing of consumer appli-
ances [43].
In summary, semi-destructive and destructive disassembly allow techniques that
are more capable of efficiently dealing with the uncertainties in product condition,
therefore allowing more economically feasible operation. On the contrary, non-
destructive disassembly tends to have high operation costs but may be unavoidable
in maintenance or for component reuse. More detail regarding the operations and
specially-developed tools can be found in Sect. 3.3.2.

2.4 Conclusions

Disassembly can be a key step in an efficient EOL treatment process, however, is

usually economically infeasible due to high operating costs relating to the vari-
ation and uncertainties in the products and process, as summarised in Table 2.1.
This chapter presents three major considerations which should be addressed to
improve the economic feasibility of disassembly.
Firstly, the disassembly plan can be optimised with respect to a goal, which can
be operating time or cost. A number of techniques regarding the representation of
the product structure and disassembly process are described. With an appropri-
ate representation, an optimal or near-optimal ordering of the disassembly opera-
tions can be found via various optimisation strategies described in Sect. 2.1.4.
Particularly the adaptive planners are of interest, since they are able to respond to
minor uncertainties like product damage. A significant amount of product knowl-
edge, e.g. the product structure or a CAD model, are required before planning.
22 2  General Disassembly Process

Secondly, regarding the completeness of disassembly, performing selective dis-

assembly to a certain depth is more feasible than the full disassembly. The optimal
disassembly depth should be determined during the planning phase.
Finally, the difficulty of the disassembly operations results from the type of fas-
teners used, and the product and fasteners’ conditions. Different fasteners can be
disestablished using different tools and techniques. Semi-destructive and destruc-
tive operations are generally preferable due to the shorter operating time and effec-
tive operation in spite of uncertainties.
In conclusion, the main source of difficulty in the disassembly process is the
need to deal with a high level of uncertainties and variations. If the disassembly is
not conducted by the product manufacturers, information regarding the product is
generally at first incomplete. Even when the expected outcomes are known, poor
product or fastener condition may require deviations to the usual plan. This causes
higher operating time in manual disassembly, and is a primary factor hindering the
industrial application of automatic disassembly.

References

1. Boothroyd G, Alting L (1992) Design for assembly and disassembly. CIRP Ann—Manuf
Technol 41(2):625–636
2. Chiodo J (2005) Design for disassembly guidelines. Active Disassembly Res. http://www.acti
vedisassembly.com/downloads/ADR_050202_DFD-guidelines.pdf
3. Duflou JR, Seliger G, Kara S, Umeda Y, Ometto A, Willems B (2008) Efficiency and feasibil-
ity of product disassembly: a case-based study. CIRP Ann—Manuf Technol 57(2):583–600
4. Gungor A, Gupta SM (1998) Disassembly sequence planning for products with defective
parts in product recovery. Comput Ind Eng 35(1–4):161–164
5. Lambert AJD (2003) Disassembly sequencing: a survey. Int J Prod Res 41(16):3721–3759
6. Kroll E, Beardsley B, Parulian A (1996) A methodology to evaluate ease of disassembly for
product recycling. IIE Trans (Institute of Industrial Engineers) 28(10):837–845
7. Mok HS, Kim HJ, Moon KS (1997) Disassemblability of mechanical parts in automobiles
for recycling. Comput Ind Eng 33(3–4):621–624
8. Gupta M, McLean CR (1996) Disassembly of products. Paper presented at the 19th inter-
national conference on computers and industrial and engineering, Computer Industrial
Engineering
9. Lambert AJD, Gupta M (2005) Disassembly modeling for assembly, maintenance, reuse, and
recycling. CRC Press, Boca Raton, Fla.
10. Gungor A, Gupta SM (2002) Disassembly line in product recovery. Ann Rev Control
40(11):2567–2589
11. Tumkor S, Senol G (2007) Disassembly precedence graph generation. In: SAM 2007—IEEE
international symposium on assembly and manufacturing, pp 70–75
12. Bourjault A (1984) Contribution à une approche méthodologique de l’assemblage automa-
tisé: elaboration automatique des séquences opératoires (Contribution to a systematic
approach of automatic assembly: The automatic generation of operation sequences).
Université de Franche Comté, Besançon, France
13. Fazio D, Whitney TL (1987) Simplified generation of all mechanical assembly sequences.
IEEE J Robot Autom RA-3 (6):640–658
14. Homem De Mello LS, Sanderson AC (1990) AND/OR graph representation of assembly
plans. IEEE Trans Robot Autom 6(2):188–189
References 23

15. Woller JD (1992) A combinatorial analysis of enumerative data structures for assembly
­planning. J Des Manuf 2(2):93–104
16. Kara S, Pornprasitpol P, Kaebernick H (2006) Selective disassembly sequencing: a methodology
for the disassembly of end-of-life products. CIRP Ann—Manuf Technol 55(1):37–40
17. Lambert AJD (1999) Linear programming in disassembly/clustering sequence generation.
Comput Ind Eng 36(4):723–738
18. Martinez M, Pham V-H, Favrel J (1997) Dynamic generation of disassembly sequences. In:
IEEE symposium on emerging technologies and factory automation, ETFA, pp 177–182
19. Zussman E, Zhou M, Caudill R (1998) Disassembly Petri net approach to modeling and
planning disassembly processes of electronic products. In: IEEE international symposium on
electronics and the environment, pp. 331–336
20. Wang Y, Li F, Li J, Chen J, Jiang F, Wang W (2006) Hybrid graph disassembly model
and sequence planning for product maintenance. Paper presented at the IET conference
publications
21. Gungor A, Gupta SM (1997) An evaluation methodology for disassembly processes. Comput
Ind Eng 33(1–2):329–332
22. Lambert AJD, Gupta SM (2008) Methods for optimum and near optimum disassembly
sequencing. Int J Prod Res 46(11):2845–2865
23. Shan H, Li S, Huang J, Gao Z, Li W (2007) Ant colony optimization algorithm-based dis-
assembly sequence planning. In: IEEE international conference on mechatronics and
automation
24. Shih L-H, Chang Y-S, Lin Y-T (2006) Intelligent evaluation approach for electronic product
recycling via case-based reasoning. Adv Eng Inform 20(2):137–145
25. Kaebernick H, O’Shea B, Grewal SS (2000) A method for sequencing the disassembly of
products. CIRP Ann—Manuf Technol 49(1):13–16
26. Tang Y (2009) Learning-based disassembly process planner for uncertainty management.
IEEE Trans Syst Man and Cybern—Part A: Syst Humans 39(1):134–143
27. Turowski M, Tang Y, Morgan M (2005) Analysis of an adaptive fuzzy system for disassem-
bly process planning. In: IEEE international symposium on electronics and the environment,
pp 249–254
28. Grochowski DE, Tang Y (2009) A machine learning approach for optimal disassembly
­planning. Int J Comput Integr Manuf 22(4):374–383
29. Veerakamolmal P, Gupta SM (2002) A case-based reasoning approach for automating disas-
sembly process planning. J Intell Manuf 13(1):47–60
30. Gao M, Zhou MC, Tang Y (2005) Intelligent decision making in disassembly process based
on fuzzy reasoning Petri nets. IEEE Trans Syst Man Cybern B Cybern 34(5):2029–2034
31. Salomonski N, Zussman E (1999) On-line predictive model for disassembly process planning
adaptation. Robot Comput-Integr Manuf 5(3):211–220
32. Lee K-M, Bailey-Van Kuren MM (2000) Modeling and supervisory control of a disassembly
automation workcell based on blocking topology. IEEE Trans Robot Autom 16(1):67–77
33. ElSayed A, Kongar E, Gupta SM, Sobh T (2012) A robotic-driven disassembly sequence
generator for end-of-life electronic products. J Int Robot Syst: Theory Appl 68(1):43–52
34. Feldmann K, Trautner S, Meedt O (1996) Innovative disassembly strategies based on flexible
partial destructive tools. Ann Rev Control 23:159–164
35. Desai A, Mital A (2003) Evaluation of disassemblability to enable design for disassembly in
mass production. Int J Ind Ergon 32(4):265–281
36. Kara S, Pornprasitpol P, Kaebernick H (2005) A selective disassembly methodology for end-
of-life products. Assembly Autom 25(2):124–134
37. Nevins JL, Whitney DE (1989) Concurrent design of products and processes: a strategy for
the next generation in manufacturing. McGraw-Hill, New York
38. Pornprasitpol P (2006) Selective disassembly for re-use of industrial products. University of
New South Wales
39. Seliger G, Keil T, Rebafka U, Stenzel A (2001) Flexible disassembly tools. In: IEEE interna-
tional symposium on electronics and the environment, pp 30–35
24 2  General Disassembly Process

40. Braunschweig A (2004) Automatic disassembly of snap-in joints in electro-mechanical

devices. In: the 4th international congress mechanical engineering technologies’04, Varna
41. Karlsson B, Järrhed J-O (2000) Recycling of electrical motors by automatic disassembly.
Meas Sci Technol 11(4):350–357
42. Reap J, Bras B (2002) Design for disassembly and the value of robotic semi-destructive dis-
assembly. In: ASME design engineering technical conference, pp 275–281
43. Uhlmann E, Spur G, Elbing F (2001) Development of flexible automatic disassembly pro-
cesses and cleaning technologies for the recycling of consumer goods. In: IEEE international
symposium on assembly and task planning, pp 442–446
http://www.springer.com/978-3-319-15182-3