Control Cabinet

Manufacturing 4.0
A study on the potential of automation and digitalisation in the manufacturing
of control cabinets and switchgears in machine and systems engineering
 Control Cabinet Manufacturing 4.0

A study on the potential of automation and digitalisation in

the manufacturing of control cabinets and switchgears in machine and
systems engineering

Philipp Tempel, Florian Eger, Armin Lechler, Alexander Verl

Institute for Control Engineering of Machine Tools

and Manufacturing Units, University of Stuttgart

30. April 2017

Institut für Steuerungstechnik

der Werkzeugmaschinen und
Fertigungseinrichtungen (ISW)

2 3
 Control Cabinet Manufacturing 4.0

Contents
Index of figures 6

Index of tables 7

1 Introduction 8

2 Design and engineering 12

2.1 A consistent engineering chain/software 12
2.2 Less diversity, more thinking in functions 14
2.3 3D engineering drawings 16

3 Production and assembly 19

3.1 Preplanning, component picking and logistics 19
3.2 Multimedia devices for information portrayal 20
3.3 Mechanical processing and assembly 22
3.4 Electrical assembly and wiring 25
3.5 Testing 28
3.6 Ergonomics and workplace design 30

4 Trends in control cabinet manufacturing 31

5 Summary 33

4 © 2017 ISW, University of Stuttgart 5

 Control Cabinet Manufacturing 4.0 Control Cabinet Manufacturing 4.0

Index of figures Index of tables

1.1 The phases studied in control cabinet manufacturing. Order creation and control 1.1 Categories for the surveyed companies in design/engineering and
cabinet/machine system start-up are not included in this study. 8 manufacturing/assembly 9
1.2 Working times from classic, standardised and automated control cabinet 1.2 Components in an average control cabinet and the average working times
manufacturing for design/engineering and production/assembly 10 per component. This determines the overall average production time for a
2.1 Phases in the design/engineering of a control cabinet and times as a proportion control cabinet 11
of overall design time 12
3.1 Phases in the production/assembly of a control cabinet and times as a proportion
of overall design time 19

6 © 2017 ISW, University of Stuttgart © 2017 ISW, University of Stuttgart 7

 Control Cabinet Manufacturing 4.0

1 Introduction
The trend towards Industry 4.0 is not a new topic in mechanical engineering. This is Table 1.1: Categories for the surveyed companies in design/engineering and production/assembly

evident both from the activities currently taking place in mechanical engineering com-
panies, and has been evident for years in academic research such as that taking place Category Design/Engineering Production/Assembly
at the University of Stuttgart’s Institute for Control Engineering of Machine Tools and
Manufacturing Units (ISW). Numerous research projects are being studied here including Classic Project/order based approach Circuit-diagram based
cloud control engineering, value-added services for mechanical engineers and users, Creation of a circuit diagram (CD) production performed page by page.
and a modern interconnected production facility, the Arena 2036. Although the ISW Schematic layout plan No prefabrication
focuses on modern manufacturing facilities for vehicles of the future, the issue is the Bills of material in Excel (manual)
same for general mechanical engineering: How to interconnect manufacturing and apply
Standardised Transition from project-based List-based manufacturing
the data it creates to the entire value chain rather than to just one manufacturing phase?
to template-based engineering Prefabrication
Design and production are two main areas within a product’s value chain where person-
Templates available NC machine
nel undergo differing training and also apply differing working practices, methods and
Creation of automated wire lists Wire sets
tools. Engineers frequently have differing expectations of the production practices that High level of repetition Partial continuous production
are eventually applied. Both design and production aspire to the same outcome but use
differing vocabulary to express this. Automated Techniques of generation, Line production
option and maximum project Processing systems
Using control cabinet manufacturing as an example, this study applies knowledge gained Manufacturers with configurators Prefabrication of all components
from research, and analyses the value chain for a control cabinet for tooling machines and Good utilisation of machinery
systems. Of the phases in the development of a tooling machine or system (see Figure
1.1), the “Order receipt/creation” and “System start-up” sub-steps are not examined in
this study. Manufacturing a control cabinet is primarily manual and faces pressure from
increasingly shorter delivery times and a high level of individualisation of tooling machines. A questionnaire was used to categorise the companies to gain an impression of control
Cabinet manufacturers need to respond quickly and flexibly to requirements and to effici- cabinet manufacturing in the branch. This served as preparation for on-site discussions
ently manage existing and future challenges: the 4th Industrial Revolution offers a multitude carried out according to the categories in Table 1.1. The categories differentiate between
of opportunities to master such challenges. The cornerstone of this study is the question “Classic”, “Standardised”, and “Automated” control cabinet manufacturing. These
“How will control cabinets be manufactured in the future?”. Companies from different areas classifications were applied to the areas of design/engineering as well as to production/
of mechanical engineering have participated in the study providing a broad cross-section assembly. The companies were also asked to provide statistics on their control cabinet
of German engineering businesses. Small machine and systems manufacturers with an manufacturing to enable comparisons to be made based on an average control cabinet
annual output of only several dozen control cabinets and systems, as well as major, well- as shown in Table 1.2. Such quantification enables an objective, quantitative response to
known companies took part in the study and provided us with insights into their processes the study questions rather than subjective impressions. The distribution of working times
of control cabinet manufacture. In total, 12 companies took part in the study. in both areas of control cabinet manufacturing as can be seen in figures 2.1 and 3.1.

Figure 1.1: The phases studied in control cabinet manufacturing. Order creation and control cabinet/machine
system start-up are not included in this study

8 © 2017 ISW, University of Stuttgart © 2017 ISW, University of Stuttgart 9

 Figures 1.2: Working times from classic, standardised and automated control cabinet Table 1.2: Components in an average control cabinet and the average working times per component.
manufacturing for design/engineering and production/assembly This determines the overall average production time for a control cabinet

Component (average) Work unit (WU) Time/WU min

Holes (drilling in body) 10 5

Cut-outs (body) 4 10
Cable ducts (sections) 20 1.5
Holes/Threads mounting panel 100 2
Installing devices 30 2
Labelling (devices) 200 2
Wiring connections 500 4.5
Terminal construction plans 130 1
Testing 1 240

Production time (total in h) 57 h

From these figures, time-savings between the categories can be determined which can
be further improved through exploiting potentials. For example, companies which still
work on a project/order basis can make time savings of up to 45% in their engineering
processes by shifting to function-based, modular circuit diagrams. In addition to pro-
posing several pragmatic measures to achieve improvements, the study also engages
with the use of 3D production data. What potential and opportunities are offered by 3D
engineering drawings and are they the key to the digital factory? The study also looks at
working approaches in production facilities and contrasts these to car manufacturing.
It is well-known that car manufacturing applies continuous production; despite a high
product individualisation, a high degree of component standardisation has still been
achieved. Is there a similar approach for control cabinet and switchgear manufacturing?

This study is structured as follows: Section 2 engages with the potentials in design and
engineering. Section 3 deals intensively with the area of production and assembly. Both
sections describe individual areas of activity, the problems these have, possible solu-
tions and opportunities. Section 4 reports on trends, from the companies’ perspectives,
in control cabinet manufacturing. Section 5 concludes the study with a summary and
outlook of Control Cabinet Manufacturing 4.0 of the future.

10 © 2017 ISW, University of Stuttgart © 2017 ISW, University of Stuttgart 11

 Control Cabinet Manufacturing 4.0

2 Design and engineering

This section examines phases in the design and engineering of control cabinets. It is as- Software landscapes of the past has led to the creation of a multitude of small stand-alone
sumed that requirements for the machine, system or control cabinet are already known systems which makes a smooth transfer of CAD data to procurement, pre-planning,
which have then been used to create engineering drawings and circuit diagrams (CD) logistics and quality control systems impossible without additional actions. Although there
from which bills of material can be derived and ordering triggered. The phases that are are now integrated solutions such as the EPLAN Engineering Configuration which can
examined as well as their core timings are portrayed in Figure 2.1. The project planning process CAD/CAE designs as well as the entire product development processes, many
and circuit-diagram creation phases require more than 50% of overall time, and are companies do not use them. Just 18% of the participating companies have recently
where significant potentials for savings exist. The subsequent phases of CD checking begun using the EPLAN tool to manage new machinery and systems series more quickly
and documentation require almost a third of the overall working time. As this section will and configure more simply, also by making using an integrated configurator. The remaining
reveal, these phases contain optimisation potential for time and financial savings. 82% of companies use highly diverse software systems for their CAD, CAE, ERP etc.
tasks. Many use the well-known SolidWorks CAD software, but in combination with
different ERP systems such as from abas, SAP, Infor or BAAN. For CAE design, EPLAN
2.1 A consistent engineering chain/software Electric P8 in combination with EPLAN Pro Panel are used with only few exceptions.

Current-day product development processes (PDP) are usually possible only through The use of separate, specialist software systems creates the problem of data isolation.
the intensive utilisation of software-based systems to perform tasks. This is the reason File formats and stored files from, e.g. SolidWorks, cannot be imported directly into, e.g.
why a multitude of different software systems exist: good software is usually restricted SAP and then processed without further ado. Most companies have therefore created
to specific rather than general applications. For example, CAD systems are available for their own data transfers solutions. For companies with a low annual output of control ca-
architects and engineers but they lack costing or ordering functions. These are provided binets, documents and data are manually transferred to the relevant software for further
by solutions such as Excel or ERP systems. Different software systems therefore need processing. 90% of companies have created their own largely automated solutions using
to be used for different phases in control cabinet manufacturing. An engineering drawing XML-file exports which can then be re-imported semi-automatically. The desire to have
created with CAD software will need to be converted into a different format before being a continuous (or at least a minimally fragmented) engineering chain means software en-
used by procurement or logistics. gineers have developed their own macros or plug-ins for the relevant software systems.
Maintaining and adapting such software plug-ins is extremely time consuming and costly
and so is not being seriously pursued by companies. Furthermore, the latest versions of
the software systems are not always being used.

This highly diversified range of software systems present many customers with obstacles
not only in engineering but also in production. Companies categorised as classic per-
form a multitude of manual phases in data transfer (e.g. the creation of bills of material or
terminal labelling). Companies who have standardised their control cabinet manufactu-
ring are already able to automatically generate the necessary documents from enginee-
ring drawings. Automated control cabinet manufacture, on the other hand, gives access
to an integrated software solution so that data no longer requires transferring from one
software system into another.

Figure 2.1: Phases in the design/engineering of a control cabinet and times as a proportion of overall time

12 © 2017 ISW, University of Stuttgart © 2017 ISW, University of Stuttgart 13

 There are three feasible approaches to overcome this issue. The first is to use a single ments of a machine (or a control cabinet/switchgear in job order/contract production).
software system which can be applied to all, or at least the main, phases in control cabinet This list is then transferred to a CAD/CAE system to create a 2D CAD engineering dra-
manufacture (these phases being CAD/CAE design, the generation of production dra- wing for precision component placement, and a CAE drawing or a circuit diagram. The
wings, prefabrication of wires and terminals). The second is to use open interfaces within documents are then used to derive detailed bills of material, i.e. including the number
the software chain to permit the required information to be simply extracted/imported. and type of terminals, contactors, etc. The obvious disadvantage with this approach is
This of course has limitations because the information required cannot/must not always be that the wheel literally has to be reinvented for each new control cabinet. This is more
recorded within a specific software system, i.e. data models are generally not designed for time-consuming and costly than working with standardised components. Furthermore,
recording extra information. For example, saving list prices for a converter is not important habitual errors can often occur using this method so that wiring always requires manual
in a CAD/ CAE drawing, while details on a device’s dimensions is largely irrelevant for bills checking.
of material. A third approach is the introduction of a standardised file format for recording
the project. CAD/CAE would then only read in and re-save data necessary for its phase. In contrast, 72% of surveyed companies are already using template-based engineering.
This would apply also to the other software systems being used by the project. Using project templates based either on a series or specific functionalities enables a
high rate of reusability. Furthermore, in engineering only the additional components
Assuming a fully integrated, project-management system actually exists for control need to be integrated – when these have already been created as functional units,
cabinet manufacturing, the question of data transfer still remains. Many companies have the integration phase should be short and error-free. 23% of the companies using
been using the same software systems for years thereby having gained vast experience standardised engineering have recently begun using a configurator for creating engi-
in using them. This accumulated knowledge is a major aid to inducting new colleagues neering drawings. The separate functional units of a machine or system are stored in
into their tasks. But it represents a not to be underestimated obstacle to implementing a database along with their corresponding interfaces and additional information. This
new software into established engineering processes. New software implementation reduces design effort to a minimum because the configurator undertakes most of the
means engineers need to invest extra effort familiarising themselves with it, and requires work. Engineers use the configurator to select a basis machine and then the functions
a data compatible transfer of ongoing and completed projects. that the customer requires. After the machine has been assembled in this way, CAD
and CAE drawings are automatically generated. The configurator undertakes control
On the other hand, not implementing new software hinders improvements to working cabinet testing, e.g. for missing components, conflicting combinations or the erroneous
procedures. An integrated software solution enables faster completion of projects and allocation of terminals and contactors.
results in fewer errors occurring than in the manual transfer of data. The surveyed com-
panies estimated efficiency gains of up to 43% in engineering. The current approach to control cabinet manufacturing is deeply rooted in the develop-
ment of a cabinet. Most engineers of today are familiar with a task-based approach in
which the requirements of a given task/project are to be fulfilled. This approach does not
2.2 Less diversity, more thinking in functions permit unconventional ways of thinking (or in this case, functional ways thinking). Work
phases can, however, be greatly simplified and speeded up by the use of functional units
Similar to the car industry in which no two finished vehicles are identical, control cabi- (FU). Functional units are to be seen from the basis of the machine or system and are
net manufacturing has a high degree of functional variance. The range of combinations reflected in a control cabinet. An example is a tool holder compressor in a machining
is obviously not as extensive as in vehicles but diversity is however extensive. 63% of centre. From an electrical engineering perspective, it requires pressure sensors inclu-
surveyed companies claimed that the proportion of special parts in their control cabi- ding I/O terminals, the I/O terminals of the compressor, and a power supply. This func-
nets exceed 50%. 27% have a low-level of variance (smaller than 20%) as well as a 50% tional unit (FU) can be designed once, the internal circuit diagram created and tested
ratio of serial parts to special parts. Just 9% have mainly serial products in their control (i.e. without having to communicate with the FU), enabling the FU to be then used in
cabinet manufacturing. engineering. Testing of these components are therefore unnecessary or only minimal,
i.e. testing is required only of the circuit diagram at the interface to the FU.
Engineering of control cabinets still takes place in 27% of companies on a project/order
basis (classic engineering). A list of devices to be installed is derived from the require-

14 © 2017 ISW, University of Stuttgart © 2017 ISW, University of Stuttgart 15

 One fact that needs be mentioned that could prevent implementation of standardised It is an evolutionary document which will undergo adaptation and changes beyond a
engineering is that design engineering training must undergo change. It requires a shift specific project period. Minor aids, such as improved virtual portrayal for better visual
from project-based thinking to modular, mechatronic based thinking. A restructuring of perception, can generate significant benefits and simplify manufacturing. Modern day
training is required not only in universities and vocational colleges but also for in-house computers are capable of portraying complex, 3D objects that are recognisable to
and further training of personnel. even untrained viewers. However, 63% of companies still use only 2D engineering dra-
wings in control cabinet manufacturing. The disadvantage to engineers is that only two
For the majority of the surveyed companies (55 %), a modular approach to engineering of the three dimensions are being used in design. Because space is limited in control
and production is not an option. The main argument against implementation of functional cabinet manufacturing, components need installing also on the side panels and doors.
development was given as the effort required for FU creation as well as the high level Without good visual aids, it is impossible to identify spatial collisions in flat drawings.
of diversity in control cabinets and machinery/systems. It was also argued that there is Furthermore, a flat representation of a control cabinet does not permit the extraction
a lack of expertise in FU creation due to the lack of clarity on how a control cabinet or of meta information such as wire lengths because the information on device height is
machine/system can be structured according to functional units. This process may, in missing to enable a determination of correct wire lengths.
certain circumstances, be rather a task for the development of a new machine series
rather than for control cabinet production. But the majority of surveyed companies still rely on 2D engineering drawings. The
reason given is the time and costs involved in conversion. In addition to 3D drawings
Companies who are already working according to functional engineering, thereby ena- offering improved visual representations and overviews, extra information can also be
bling the use of (online) configurators, report substantial time gains in control cabinet saved for use in other process phases. The required wire and cable lengths between
manufacture. Up to 44% savings in time and costs are possible. The main work in a pro- individual components and cable duct capacities can be automatically determined.
ject then concerns only the availability of the components and a quick optical inspection Dimensions can be correctly determined because the spatial arrangement of devices
of the engineering drawings (CAD and CAE). However, new areas of activity need to be can be checked beforehand thereby enabling automation of the control cabinet’s
created for machine standardisation and the creation and maintenance of FUs. But this mechanical processing. A design model can be virtually displayed during mechanical
can be achieved independently from ongoing orders and can therefore be seen as a part and electrical production, providing technicians instant visual feedback to check the
of a process of ongoing improvement. correctness of their work.

The benefits of function-based engineering are obvious. Using functional units minimises In addition to direct checking during work phases, information from a 3D control cabi-
checking effort and can be performed independently of ongoing project work. Not least net can also be used for the “digital twin”, i.e. the virtual portrayal of a control cabinet
of all, design plans derived from a standardised engineering chain-, can be assumed to that can be constructed and tested in parallel to its actual manufacturing. This virtual
be basically error-free. This eases the work of design engineers, and mechanical and digital twin can be used, prior to actual manufacturing, to virtually run the control soft-
electrical technicians can work more efficiently due to the production plans generally ware, eliminating the need for this task to be carried out at the end of manufacturing.
being correct.
This topic raises the general question of why 2D engineering plans are still being
created in engineering when manufacturing of a control cabinet takes place anyway in
2.3 3D engineering drawings three dimensions. Working in 2D is counter-intuitive and counter-productive yet is being
practised more often than not.
An engineering drawing is only as good as the information it contains and the infor-
mation that can be directly derived from it. A good engineering drawing also enables To gain the full benefits of 3D design requires companies to exploit the potential of this
information to be read and understood quickly without requiring detailed knowledge technology in multiple manufacturing phases. The most benefits are in the engineering
of the object to be created. An engineering drawing is essential to enable mechani- phase where data is created and modified. In subsequent phases, data is generally only
cal and electrical technicians to carry out error-free production tasks. However, from read. In addition to the implementation of a CAD and CAE systems with 3D drawing
a designer‘s/engineer’s perspective, an engineering drawing is much more than this. capabilities, existing data also needs to be extended and possibly migrated. The effort

16 © 2017 ISW, University of Stuttgart © 2017 ISW, University of Stuttgart 17

 Control Cabinet Manufacturing 4.0

Production and assembly 3

required to extend data (adding component heights, recording CAD models etc.) is This section examines the procedures in the production and assembly of control ca-
usually the biggest hurdle given by many companies for not migrating data. Correctly binets. It assumes that certain documents have been available from engineering that
processing CAD data can be difficult especially in the context of modularisation and have triggered component orders, the majority of which have already been delivered.
the structuring of a control cabinet into functional units. But the time gains should not The phases as well as their core timings are portrayed in Figure 3.1. Almost 75% of
be ignored – up to 35% of engineering time and up to 22% of production time can be time is required for the phases of mechanical component assembly and wiring (inclu-
saved because drawings can be fully trusted. ding electrical component assembly) where significant potential for optimisation can
be assumed. The phases before and after these account for the remaining 25% where
3D engineering drawings also offer benefits for technicians in production. These are no significant improvements can be realised. As this section will reveal, these phases
provided separately in the following sections. contain optimisation potential for time and financial savings.

3.1 Preplanning, component picking and logistics

Preplanning is a major element in control cabinet manufacturing. Control cabinets generally

comprise a multitude of terminals and wiring. Well prepared production begins with
component picking, i.e. the assembling of the components required for a specific
control cabinet. A long outdated approach is undertaken by the technician who alone
organises the required components. This results in lengthy, criss-crossing procedures.
The move towards Industry 4.0 involves isolating tasks and areas of expertise so that
savings in time and costs can be made via picking of the majority of components for a
control cabinet.

Figure 3.1: Phases in the production/assembly of a control cabinet and times as a proportion of overall time

18 © 2017 ISW, University of Stuttgart © 2017 ISW, University of Stuttgart 19

 Components in control manufacturing can be classified into three categories according ment and testing documentation. Masses of documents numbering several hundred
to their size and purpose and are therefore generally located in different areas within pages are not unusual.
production halls. Quantifiable components, such as drive amplifiers and terminals
are held in large storerooms. Wire and cable spools are located near the production A paper-based production system is not really a disadvantage, but does represent
machines where wiring and cabling takes place. At the companies visited, non-quan- a major challenge to keep order of the vast quantity of paper. For example, a circuit
tifiable components, i.e. bulk material, used different storage methods ranging from diagram is made up of an average of 320 pages with electricians needing to browse
centralised kanban systems through to mini stockholding systems at the workplace. through them several times during wiring to find wire destinations. Simple portrayal of
routing information, such as that provided by a wiring list or a digital circuit diagram,
As already mentioned, control cabinets are comprised of a multitude of components is not possible. Furthermore, paper documents are exposed to surrounding condi-
making mechanical component assembly and wiring a lengthy, manual process. This tions, e.g. the grime of a production hall can make documents unreadable. Changes
can, however, be made significantly faster by intelligent use of existing information taking place during production (e.g. component repositioning because of a collision,
in 3D models. A dedicated procedure of preplanning and commissioning enables or the re-correction of wiring details due to a triplicate terminal connection) are not
the prefabrication of terminals (positioned in strips and labelled) to make installation major problems but need to be recorded and subsequently reconciled with the digital
significantly faster and eliminating transfer routes. The same applies to wiring which ac- documentation. For such change management, the surveyed companies use two main
counts on average for almost 60% of production time. By using prefabricated wiring or approaches: changes are either marked in the relevant engineering documents (e.g.
wiring sets (at the required length and thickness and with connectors added), time sa- circuit diagrams) with a cover sheet listing all amended sections, or changes are made
vings of up to 35% can be achieved. In combination with list-based wiring, up to 50% exclusively in a detailed log/table of changes. Both approaches are based on having
is achievable. To exploit such potential, workplaces need to be adapted with boxes or one further, multi-page, detailed paper document and therefore do not help the paper
mobile shelving for storing the picked goods. Furthermore, detailed and checked bills problem. They also introduces an additional task at the final documentation phase of
of material from engineering are needed to guarantee accurate picking. having to transfer changes from handwritten notes to the digital engineering drawings.

This study makes it clear that for control cabinets of the future, components need to be A further problem with paper documents, according to all surveyed companies, is the
prefabricated and available at the relevant workplaces, and bulk materials need to be provi- persistency of changes being made to control cabinets. Comments or changes to a
ded to technicians via kanban systems, irrespective of whether nested or continuous document are easy to make using a pencil. However, major changes, e.g. customer
production systems are being used. changes requiring different components, cannot take place without further action.
Such modifications generally take place within the engineering department so it can
take time until the updated data arrives at production. According to the surveyed com-
3.2 Multimedia devices for information portrayal panies, this can take up to a week or even 2 in exceptional cases.

People today are surrounded by digital devices and multimedia information sources. During this time, further changes might have taken place further complicating change
The latest stock exchange indices, news or images of friends are just a click away. management and highlighting the problem of paper-based data processing.
Routine use of smartphones or tablets is no longer the exception. It was common in
the production halls of the surveyed companies for primarily paper-based production to The surveyed companies were not only open to the proposition of digital document
be seen as no longer appropriate. However, all surveyed companies, with few excep- folders, some had already considered doing it. Two companies had already carried out
tions, rely primarily on analogue documentation in production. initial evaluations on the feasibility and effort required to implement digital document
folders. Overall, many engineers and production supervisors as well as mechanical and
Preplanning digitally access component and stock lists and then print them out for the electrical technicians would welcome digital document folders to enable faster viewing
picking trolleys. In the actual production of the control cabinet, i.e. mechanical and of engineering drawings as well as the relevant production information. They see major
electrical processing and assembly, only printed documents and handwritten notes are benefits in this respect. For example, such information can be displayed relatively easily
being used in the form of circuit diagrams and wiring lists through to change manage- for wiring. The average time for wiring using a circuit diagram is 54 h with around 31%

20 © 2017 ISW, University of Stuttgart © 2017 ISW, University of Stuttgart 21

 of the time required for preparatory tasks such as reading the diagram and locating Mechanical processing of a control cabinet includes processing of the body via drilling
source and destination points. A further 13% of time is required for preparatory work and the creation of recesses for switches, buttons, cooling units and cut-outs, as well as
on the wire – initial routing and estimating wire lengths. The remaining 56% is for the the processing of mounting panels (none of the surveyed companies use rail systems in
actual wiring – cutting, adding connectors, crimping. A third of the time is therefore their control cabinets). Only 45% of companies carry out this phase themselves. A small
spent reading the document – with an average of 500 wires per control cabinet and an majority obtain their control cabinets prefabricated from a supplier. Although this has
average time of 54 hours for wiring, this equates to 16.74 hours per control cabinet. A the advantage of eliminating the need for a mechanical processing department – which,
software-based system which can visually portray the source and destination points for low annual outputs of control cabinets, would be underutilised and have a high rate
of a wire – whether in a digital circuit diagram or a virtual 3D CAD/CAE drawing – can of downtime – control cabinet manufacturing then becomes dependent on supplier
reduce this effort by up to 81%. delivery terms. Companies with low annual outputs of switch-gears and cabinets lose
competitiveness because under certain circumstances the terms and conditions offe-
Digital devices can also be profitably applied in other areas of control cabinet manufac- red by suppliers are not competitive.
turing. In electrical component can be visually displayed assembly, information on the
location of a component by highlighting it in a CAD drawing. To use such new media The surveyed companies with in-house mechanical processing perform it exclusively
requires consistent and accurate engineering drawings. Information to be displayed in with automated processing machines – either punching machines or CNC milling/
production must be created in a preceding phase. This requires at the least a detailed drilling machines. Apart from different processing techniques, the main differences
and tested 2D engineering drawing – whether this has been created using maximum between these companies concern the use of engineering drawings in the processing
techniques or configurators is of less importance. In addition to the generation of accu- procedure. In 58% of cases, machine instructions are performed by manually inputting
rate production data, production availability also needs to be achieved. This is dimensions for cut-outs and drilling. This creates susceptibility to incorrect input and
generally an information technology task which can be undertaken by internal IT de- requires accurate reading of the engineering drawing. High precision is required here
partments or outsourced. because drill holes for fixing components generally involve low production tolerances –
screw positions for voltage inverters are specified by the manufacturer. The remaining
Depending on the existing production halls, wireless communication is possible but a 42% have mainly automated processing. The loading and unloading of CNC machines
cable-linked system would also be sufficient. Providers of such software systems need takes place manually but processing profiles already exist in the machines and require
to develop detailed security systems with role-based permissions to prevent unautho- only correct loading by the technician. These profiles can be taken from 2D or 3D CAD
rised changes taking place. This might be achieved using an approvals function at final drawings or can be based on standardised components– e.g. just one specific type of
testing of the control cabinet. Furthermore, issues such as what to do with contradic- kill switch for ISO certified control cabinets.
tory or simultaneous changes need to be addressed as well as the issuing of change
notifications. In addition to the mechanical processing of a control cabinet body and mounting panel,
the cutting to length of DIN rails and cable ducts can also be standardised and automa-
ted. Surprisingly, all of the surveyed companies perform manual cutting and installation
3.3 Mechanical processing and assembly of components based on an evaluated engineering drawing. Minor mechanical
aids are used to simplify cutting but details on the length and diameter of cable ducts
Mechanical production comprises mechanical processing and mechanical assembly are taking from engineering drawings – no list of the lengths to be created exists similar
and is the first of two basic phases to produce a functioning control cabinet from engi- to a wiring list.
neering drawings. Among the surveyed companies, both phases are mainly performed
manually so it can be assumed that major potential exists for digitalised and (semi)-
automated production. This section looks more closely at the phases “mechanical
processing” and “mechanical assembly”.

3.3.1 Mechanical processing

22 © 2017 ISW, University of Stuttgart © 2017 ISW, University of Stuttgart 23

 It is easy to see that among the multitude of these manual tasks errors can occur production, up to 47% is achievable. In addition to these possible time savings, automa-
which could be avoided by a few easily realisable steps. As a rule, the engineering ted control cabinet manufacturing can also save on material costs. Digital engineering
drawings for a control cabinet exist prior to production beginning so they can be used drawings can be used to optimally obtain the required lengths of cable ducts from exis-
to obtain production information needed for mechanical processing. By implementing ting stocks thereby reducing or possibly even preventing wastage. This provides both
this step in the engineering of a control cabinet, the mounting panel could be clamped economic and ecological benefits. However, such automated phases require consis-
in the relevant processing machine and automatically provided with the required drill tent and checked 3D engineering drawings of the control cabinet.
holes and cut-outs. This approach could also be applied to the walls of a control cabi-
net body as these are generally detachable from the cabinet frame. Up to 55% of time
savings in mechanical processing can then be achieved. 3.4 Electrical assembly and wiring

The use of a CNC mini/drilling machine (manual or automated programming) is an All participating companies perform the assembly and wiring of the cabinet in-house.
option for some of the surveyed companies (27 %) due to the purchasing and opera- 27% of companies still acquire their control cabinets pre-fabricated, i.e. already with
ting costs exceeding capacities. This issue is mainly academic for companies with low cut-outs and recesses. The proportion of mechanical assembly and wiring therefore
annual manufacturing output because the high rate of utilisation necessary for these takes up almost 75% of processing time (see Figure 3.1). Electrical assembly and wiring
machines cannot be provided. On the other hand, obvious benefits are provided by is the phase in control cabinet manufacturing that is capable by far of benefiting from
CNC-based processing such as a high level of absolute and repeat precision. digitalisation and automation. “Electrical assembly” and “electrical wiring” are examined
separately in the following.

3.3.2 Mechanical assembly

3.4.1 Electrical assembly
Following mechanical processing of a control cabinet (mounting panel, side walls and
doors), DIN rails, cable ducts, seals and brackets for the components are installed. The electrical assembly of a control cabinet is the first of two steps in electrical produc-
For all surveyed companies (only 45% mechanically process their control cabinets in- tion in which devices (transducers, contactors, signalling devices, power supplies etc.)
house), this phase takes place manually with mechanical aids being used only for cut- as well as smaller components such as terminals, switches and buttons are installed.
ting the DIN rails and cable ducts to length. The production instructions are obtained As a basis for assembly, technicians use either bills of material (for 45% of the surveyed
by technicians from the available engineering drawings and transferred to the process. companies) or circuit diagrams (55 % of the surveyed companies). Assembly using bills
of material is beneficial because such bills are pre-sorted by component thereby giving
The required small parts (screws, nuts, sleeves etc.) are generally acquired by techni- an easy overview. Using a circuit diagram to determine components requires much
cians themselves in a component picking phase. In any event, there is no packaging of more effort and is more prone to errors with several hundreds of pages needing to be
the required bulk parts although this would bring benefits by avoiding transfer routes browsed. 81% of companies do not pick bulk goods (terminals, screws, jumpers) due
and incorrect estimations of the required quantities. However, much more serious is to it being impractical. Technicians are therefore required to maintain a certain amount
the non-utilisation of existing information from engineering drawings for automated of bulk goods at their workplaces. 22% of companies use a kanban systems while all
production. other cases require technicians to take responsibility for monitoring stocks themselves
and, when necessary, replenishing them from a centralised store. The remaining 18%
Engineering drawings today can be provided with additional information which can also of surveyed companies obtain their terminal strips prefabricated from suppliers thereby
give production instructions. Engineering drawings exist prior to production starting saving both picking and terminal-strip fabrication. Installing such terminal strips redu-
so in principle it is possible to use the drawings to obtain the lengths of DIN rails and ces effort by up to 92%.
cable ducts for automated cutting. It is even feasible to use a processing centre where
cutting is carried out as well as the fixing or riveting of DIN rails and cable ducts. In fully
automated production, this can enable up to 83% time savings, and in semi-automated

24 © 2017 ISW, University of Stuttgart © 2017 ISW, University of Stuttgart 25

 The installation of devices can take place only when a relevant engineering drawing mation for wiring and can be considered as a complex portrayal of a list of links. Circuit
is available. Such drawings provide the positioning of cable ducts and DIN rails, with diagrams are also difficult to work with because of their counter-intuitiveness and non-
specific areas then being designated for inverters, bus couplers and terminals. In 18% linearity: components that in a control cabinet are placed sequentially to each other will
of cases, control cabinets are assembled without engineering drawings and use only not necessarily be described on sequential pages in the corresponding circuit diagram. It
circuit diagrams as a basis. Reasons given for this are too shorter project times and is more the rule than the exception that technicians need to browse through the com-
therefore no option for preplanning and the corresponding engineering. plete circuit diagram multiple times during the wiring process to connect all components
together. In addition to the average 42 seconds required for wire routing, a further 67
As in other phases of control cabinet manufacturing, automated assembly using terminals seconds are needed for reading the circuit diagram and locating the components on the
or devices requires a detailed model of the control cabinet. In the majority of surveyed mounting panel and/or side walls. Another wiring task is wire preparation, i.e. cutting to
companies (72%), the number of terminals are manually taken from the circuit diagram. In- length, crimping and labelling, which requires another 157 seconds. In total, an average
creased effort is therefore required due to the task being manual and error prone. This can technician requires an average of 266 second to wire together two control cabinet com-
be avoided and made faster by taking relevant action in the engineering phase. A digital ponents – this represents a good 4 1/2 minutes per wire.
model of a control cabinet can also be used for terminal prefabrication by in-house prepl-
anning or an external service provider. This delivers time gains of up to 90% and stockhol- To perform wiring using a wiring list is a more advanced approach than classic wiring
dings can be reduced via a just-in-time supply of fabricated terminal strips. using a circuit diagram. The wiring list method is being used by only 9% of the sur-
veyed companies. Even in these cases, control cabinets are usually wired according
If standardised components or “functional units” are used in the engineering of control to circuit diagrams due to technicians finding them easier to read. Interestingly, this is
cabinets, terminal packages or complete terminal strips can, due to standardisation, exactly the opposite to what a wiring list should achieve. The information in a wiring
be kept in stock as prefabricated parts. These can then be supplied in preplanning list includes, in addition to source and destination points, additional information such
and/or picking, together with the functional units. If this approach is combined with as wire length, diameter, colour and labelling. Further information is also available as
the outsourcing of terminal prefabrication to external service providers, up to 90% required. A wiring list allows a much more linear realisation of the wiring then a circuit
time-savings can be gained because terminal-strip bundles are capable of high-speed diagram. To create a complete and correct wiring list requires, however, additional
installation using special clip systems. information from engineering such as the type of wire as well as its thickness – this in-
formation depends on the component to be wired and therefore needs to be recorded
in the CAD-/CAE models.
3.4.2 Electrical wiring
In 72% of the surveyed companies, errors are usually first detected at the production
A main phase in control cabinet manufacturing by far is electrical wiring in which de- stage because drawings are not checked in engineering. Such errors include overfull
vices are connected to each other to produce an electrically functioning control cabi- cable ducts, collisions between devices and body components, and hindrances to
net. This phase is also by far the most time-consuming phase for all surveyed compa- thermal convection. Electrical technicians are trained well enough to enable them to
nies, requiring 49% of overall production time irrespective of the quality of information rectify such errors and to record the changes to a change log. At the end of control
coming from engineering. The technicians performing the work need to be highly cabinet testing, the information in the logs is transferred to the digital documentation
trained to read circuit diagrams and have the skills to spot possible errors. In 90% of to ensure it matches the end product. This step would be unnecessary in fully digita-
companies, the circuit diagrams is used as a basis for wiring which provides informati- lised production because technicians would enter the changes directly into the digital
on only on the source and destination of a connection. Additional information such as engineering drawings.
wire length, colour, thickness and cable-end configuration would enable faster wiring.
To revolutionise wiring, different solutions and approaches need to be considered. If
The correct selection of wire thickness is important for correct cabinet functioning circuit diagrams are being used, time savings of up to 32% can be achieved through
and depends on the technician being highly trained. Circuit diagrams, which frequently the structured portrayal in a wiring list of the procedure, especially the actual wiring.
comprise hundreds of pages, are not very informative, generally lack important infor- Furthermore, wiring using a wiring list does not require the skills of highly qualified

26 © 2017 ISW, University of Stuttgart © 2017 ISW, University of Stuttgart 27

 technicians, which represents an economic benefit when requiring extra staff in peak correct functioning of individual components. Functional testing of a control cabinet
periods. In addition to the reading of wiring instructions, cutting wires to length is also a additionally tests performance thereby also validating safety circuits. Of the 75% of
time-consuming process which usually requires years of experience until it becomes fast companies performing electrotechnical testing, 16% also carry out functional testing
and virtually error-free. A circuit diagram contains no information about the routing during engineering in a virtual start-up procedure. Errors in control software can
of a wire. It is therefore impossible to use fabricated wires so technicians need to therefore be rectified prior to production. Final electrotechnical testing takes place in
perform multiple tasks: the laying of wires, preparation, completion and installation. 3D all cases by checking electrical potentials using potentiometers. The testing methods
engineering drawings, however, provide information on wire routing and length. Fur- differ only by the equipment being used (standard or programmable potentiometers).
thermore, the full-capacity of cable ducts can already be determined during enginee- Functional testing of control cabinets and switchgears use test controller equipment.
ring. This is all possible in just a few clicks using relevant 3D development software. In all cases, these are controllers, programmed by the company and created from a
corresponding machine’s “maximum project”. This enables self testing of production
As well as the already mentioned approaches, some control cabinet manufacturers are tasks as well as components. 18% of cases also perform load testing on the cabinets.
considering digital production systems in which wiring takes place without any human This requires additional preparatory effort but enables problem-free connection in the
input. The market currently has only a few providers pursuing such systems. However, field to be performed. If control cabinet testing is performed only when the cabinet is
their presence alone shows the direction in which control cabinet manufacturing is go- located in the field (18% of companies perform field testing or leave testing to the cus-
ing. The use of such systems is perhaps debatable for companies with a low output of tomer), damage can occur to the machine or system resulting in expensive repairs.
more individualised control cabinets, but those with high outputs can achieve time sa-
vings in wiring of up to 63%. It must be observed, however, that these machines are far Unfortunately there is no general solution to control cabinet testing due to the multitude of
from working as intelligently and experienced as trained technicians. Minor deviations possible configurations. Furthermore, there are also issues of different customer require-
in mechanical or electrical assembly can result in a component being incorrectly con- ments and delivery conditions, e.g. to foreign locations. For the series-type manufactur-
nected making a control cabinet inoperable. To use wiring robots or wiring machines ing of control cabinets, a general routine can be created for functional testing which
requires product development processes to be applied to the control cabinet from the can then be extended for specific modules. Such test procedures can be carried out
very start to deliver high-quality engineering and production data. Only in this way can either manually or automatically whereby an average testing time of at least four hours
the benefits be enjoyed of semi or fully automated production. (providing no errors exist) makes a semi or fully automated system (e.g. by a test robot)
preferable. Testing can then take place overnight with the results being analysed the
next morning and erroneous systems then subjected to closer examination. Unnec-
3.5 Testing essary production downtimes are then avoided.

Testing of a control cabinet is necessary for all control cabinet manufacturers to ensure A further option to achieve continuous testing in control cabinet manufacturing is
the correct production and functionality of all components. Testing does not just cover through augmented reality plus continuous process monitoring in the production phas-
final testing of a control cabinet but also ongoing tests during production. This includes es. Augmented reality merges the real and virtual worlds and enables technicians to
checking of production drawings prior to the different phases as well as optical inspec- obtain information, for example, through eyeglasses or a monitor positioned above the
tion following conclusion of a phase. 36% of surveyed companies performed intensive assembly surface. Any information can be displayed, such as details on drill holes or a
testing during phases which are performed by the technician doing the work and then component, as well the start and destination points in wiring. Continuous process
checked for correctness by the next technician. This approach requires extra effort monitoring results in the work undergoing permanent testing to enable early and au-
during the individual phase but has the benefit of identifying and rectifying errors at an tomatic detection of errors. This requires detailed and checked engineering drawings
early stage. Furthermore, each technician indirectly becomes his or her own tester and which have at least a 2D model of the control cabinet. A circuit diagram is insufficient
therefore carries out his/her work more reliably. here to identify incorrect drilling, components and wiring.

25% of surveyed companies carry out electrotechnical final testing including functional The benefits of continuous visual inspection of performed tests are an argument for
testing following wiring – the remaining 75% carry out only electrotechnical testing. implementing this approach. Any error (e.g. missing components or incorrect wiring)
Electrotechnical testing checks only correct wiring of the control cabinet and not the can then be identified and rectified before the phase is concluded. This is less time-

28 © 2017 ISW, University of Stuttgart © 2017 ISW, University of Stuttgart 29

 Control Cabinet Manufacturing 4.0

Trends in control cabinet 4

manufacturing
consuming and costly than identifying and rectifying irregularities during the production
phase. For example, to rectify drill holes that are too small and identified only during
electrical assembly require more effort because components may need to be removed As part of the study, companies were also asked about the trends they predict in engi-
prior to re-drilling. Visual inspection at the end of a phase by the technician himself/ neering and production. These are briefly presented in this concluding section.
herself makes the technician his/her own inspector and thereby promotes a more ca-
reful approach to the task. Companies which apply this approach report a reduction in
errors of up to 85% and a subsequent rise in efficiency of at least 37%. Components with greater complexity and more software

Today’s components have an increasingly larger share of software which needs to be ac-
3.6 Ergonomics and workplace design counted for primarily within engineering. The positioning of these components is generally
no longer restricted and they can be installed anywhere required within the control cabinet.
On the basis of on-site analyses, a number of improvements were identified in respect However, this results in the need for intensive software testing of the control cabinet.
of ergonomics and workplace design. During wiring, technicians often need to rotate
by 180° from viewing the circuit diagram to wiring the control cabinet. This is not only
time-consuming but also damaging to health. A physically and mentally balanced A greater number of field components through decentralisation
workload is required to promote health and achieve optimal performance. As already
mentioned, the assistance of multimedia user devices is one option to simplify wiring Today’s machines and systems are so complex and are equipped with so many func-
and ease the work of the technician by optimising the workplace. Time savings can tions that there is a diminishing need for the manufacturing of large control cabinets.
also be gained. The trend is towards decentralised, field-located control cabinets containing the neces-
sary components and a minimum of control logic. A central control cabinet will conti-
The study identified different approaches on how a control cabinet is positioned to nue to provide primary control cycles but will have a much reduced level of complexity.
perform assembly. This takes place mostly with the cabinet in a horizontal or vertical Such decentralisation has a direct affect on the complexity, density and quantity of
position and rarely at an angled position. Some components, such as drive amplifiers, control cabinets that need to be installed for a machine or system.
are heavy. The positioning of the main panel and/or the cabinet plays an important role.
The study made a detailed analysis of working procedures in production and determi-
ned that technicians can assemble components faster and more ergonomically when Restrictions on component diversity
the cabinet is in a horizontal rather than a vertical position. When positioning heavy
components, lifting aids can be beneficial. Wiring on the other hand, requires the ca- Although the majority of participants in the study referred to a large diversity of compo-
binet to be in a vertical position. Wiring is then much easier for the technician, whether nents (63% have a more than 50% share of specialist components), many also wished
sitting or standing. for a reduction in such diversity. The reason for this is, on the one hand, the effort requi-
red in administration and the multitude of specialist parts required in engineering. On the
In addition to the positioning of the control cabinet, the tools and equipment at the other hand, a high level of diversity affects production by requiring highly trained technici-
workplace also play major roles in optimising production. Tools, wires and the required ans capable of managing the corresponding high rate of production variance. Therefore,
bulk material need to be sorted and within easy reach of the technician. 11% of the 36% of companies are pursuing a reduction in component diversity by using only one or
companies use kanban systems at the workplace which provide materials automa- two variants, e.g. pressure generators. This of course requires a new way of thinking and
tically. This ideal workplace design avoids long transfer routes to centrally located may result in oversized systems and components, but will reduce effort both in enginee-
stores. For the other 89% of firms, the technicians themselves are responsible for ring and in production.
replenishing bulk-goods containers. Time savings of up to 35% are achievable by
exploiting aforementioned potentials.

30 © 2017 ISW, University of Stuttgart © 2017 ISW, University of Stuttgart 31

 Control Cabinet Manufacturing 4.0

Summary 5

Outsourcing This section briefly summarises the findings of the study. More detailed explanations can
be found in the relevant sections.
It was important for the study to survey companies who perform in-house control cabinet
manufacturing. This was the case for only 78% of the surveyed companies. This means
that every fifth control cabinet for a machine or system is already being manufactured by Automated production using 3D engineering
external providers. It should also be mentioned that 27% of surveyed companies do not
carry out their own mechanical processing on control cabinets but obtain these prefab- A clearly identifiable trend among the surveyed companies is towards the creation of
ricated. Time and cost pressures mean some companies – especially those with a low 3D engineering drawings. This may involve complications and flat learning curves in
annual output of control cabinets and switchgears – are asking themselves whether to the transitional phase from 2D drawings and circuit diagrams, but clear benefits can be
have their control cabinets manufactured externally. Other companies want to continue achieved by using 3D models. Ultimately, 3D models enable the fully automated, conti-
manufacturing their cabinets but using, when possible, prefabricated terminal strips and nuous production of control cabinets – or corresponding semi-automated control cabinet
wire harnesses. The extent to which the trend towards outsourcing continues is difficult manufacturing.
to predict. However, the attractive conditions from external providers are not to be unde-
restimated.
  Outsourcing or in-house production?

For companies with a low annual output of control cabinets with relatively high and
non-profitable development and production activities, there is the question of whether to
outsource control cabinet manufacturing. Some of the surveyed companies consider this
not only a loss of in-house skills but also a dependency on the delivery terms of contracted
manufacturers. This is generally not a problem in the series manufacturing of machines
because time is less restricted compared with the manufacturing of systems. There are
no statistics available to evaluate this so a separate analysis is required.

Standardisation in engineering

To ensure in-house control cabinet manufacturing remains competitive in the coming

years, production approaches require increased standardisation and unification. In view
of the high level of specialist parts in mechanical engineering, this may at first sight seem
impossible. However, the majority of these special parts are simply options based on a
single type of specific machine. It should therefore be possible to create a list of pro-
duction instructions for virtually any machine. This will significantly benefit production
because such lists would leave practically no room for differences in interpretation. Errors
can then be avoided-, and, in peak periods, untrained operators can be quickly deployed.

32 © 2017 ISW, University of Stuttgart © 2017 ISW, University of Stuttgart 33

 Thinking in functions

Similar to the engineering of tooling machines or systems, control cabinet manufactur-

ing must also be considered in functions or functional units. Today’s engineers, however,
generally design their control cabinets from bills of material. In the long term this is im-
practical and requires a rethinking of engineering training and the use of software. A control
cabinet needs to be developed in functional units whereby each unit can be independently
tested and managed. This enables significant potentials in engineering and production.

List-based production documentation

Documentation used in production is generally based on bills of materials, circuit dia-

grams and engineering drawings. It is not easy to extract information from these to com-
plete a specific phase and depends on the ability of each individual technician. By using
list-based production documentation which is clearly structured and sequenced, produc-
tion instructions can be carried out faster and with fewer errors. Capacities can then be
easily increased during peak periods because highly trained staff are no longer required.
List-based production is the key to error-free, faster and more flexible production.

Preplanning and component picking for faster production processes

To enable continuous control cabinet manufacturing in the future requires more attention
to preplanning. Pre-picked boxes of components for each individual control cabinet need
to be assembled by logistics in the component stores. With the exception of procured
wires and the associated bulk material, technicians can rely on having the correct quan-
tity of required components. The need for technicians to source for themselves the com-
ponents needed for a control cabinet must become something that belongs to the past.

34 © 2017 ISW, University of Stuttgart 35

 Process consulting
ƒƒ
Engineering software
ƒƒ
Implementation
ƒƒ
Global support
ƒƒ

11.2017

EPLAN Software & Service UK

Braithwell Way · Hellaby Industrial Estate · Hellaby ·
Rotherham · South Yorkshire · S66 8QY ·
Phone: +44(0)1709 704100 · Email: info@eplan.co.uk ·
www.eplan.co.uk

PROZESSBERATUNG ENGINEERING-SOFTWARE IMPLEMENTIERUNG GLOBAL SUPPORT