Throughput time reduction
of an
administrative business process
From a service providing maintenance organization
Name: Martin Veninga
II
Preface
I have written this master thesis to complete the master program Technology Management at the "Rijksuniversiteit Groningen". For a period of more than six month, I have been studying problems present at the Integrity Services & Solutions business line of Stork Technical Services (STS) Delfzijl. I have gained a lot of practical experience in a work environment that is closely related to my master program. Hereby I would like to thank everyone within STS Delfzijl for a pleasant work experience, and for their open and helpful attitude during my research period. Especially the planners and foremen of the IS&S business line who have great contribution to the realization of the solution of the present problems. In addition, I would like to thank E. van den Dool, who has offered me the opportunity to graduate at a very nice organization.
After a difficult period I can finally close a dark chapter in my personal life, by finishing my master program. I want to thank the people who have never stopped believing in me and have always supported me. I would like to thank dr. W.M.C. van Wezel for his patience, guidance and ideas during this long and hard period. From the "Rijksuniversiteit Groningen" I also want to thank dr. ir. W.H.M. Alsem as co-assessor. Finally I want to thank Dirk Messemaker, who supported me in finishing my master thesis.
III
MANAGEMENT SUMMARY
This research was conducted at Stork Technical Services (STS) located at Delfzijl in the Netherlands. It is a service providing organisation, which is dedicated to providing seamless performance to corporations and asset owners in the Chemical, Oil & Gas and Power industries as well as railway and food & pharmaceutics industries. The research took place at the Integrity Solutions & Services (IS&S) business line, which is responsible for the maintenance of pumps and valves and providing On Site Machining (OSM) & Bolting services. Management expected that the administrative business process of IS&S Delfzijl faced efficiency problems. A initial investigation showed that management was not satisfied about the efficiency of the administrative business process, which was reflected by a long throughput time of the billing process. To solve this business problem the following research question was formulated:
“How can the throughput time of the administrative business process at Integrity Solutions & Services
Delfzijl be decreased?”
At the start of the diagnosis it was validated whether the long throughput time is actual a problem IS&S Delfzijl is facing. Orders remained mainly at workstations of the administrative business process after the requested service was delivered, but still had to be administratively completed. The further course of the problem diagnosis consisted of multiple research methods, to identify and validate causes of the long throughput time. The three causes that influenced the efficiency (expressed in time) could be validated: “an unclear administrative business process”, “a bad order intake”, and “the
usage of different IT systems”. The administrative business process was unstructured and not
transparent, resulting in an inefficient process. In addition, the tasks and responsibilities were poorly defined, what resulted in a lack of discipline and sense of responsibility among members of the IS&S business line. The information obtained during the order intake is often incomplete and inaccurate, causing additional contact moments with the customer. The use of two independent ERP systems to support the execution of the administrative tasks induces many transactions and delays. A lot of precious time is lost by performing wasteful activities. All together this makes the administrative business process inefficient where the management and infrastructure of the information process is not optimal.
By improving the administrative procedure of processing an order, and the corresponding tools, it is expected that the throughput time is significantly reduced. To improve the management of information process and infrastructure it was decided to redesign the administrative business process. Literature is used to find a quality oriented theory that focuses on the improvement of throughput times of (administrative) business processes. The lean concept came up as most suitable technique to redesign the administrative business process. The lean concept is a systematic approach to identify and eliminate waste (non-value-added activities) through continues improvement by flowing the product at the pull of the customer in pursuit of perfection (NITS, 2000). The concept focuses on the removal of waste, which improves the throughput time of the business process. Eventually a redesign was developed consisting of three sub-solutions:
IV 2. The introduction of an order intake checklist: By using the order intake checklist a more
efficient procedure of accepting and processing incoming orders is created.
3. The implementation of the SAP system: By making use of one all-embracing ERP system to perform all administrative tasks instead of two independent ERP systems, delays and unnecessary transfers are declined.
V
TABLE OF CONTENT
PREFACE II
MANAGEMENT SUMMARY III
LIST WITH ABBREVIATIONS VII
CHAPTER 1. INTRODUCTION 1
1.1 STORK B.V. 1
1.2 HIGH-OVER BUSINESS PROCESS 2
1.3 BUSINESS LINES 3
1.4 PROBLEM IDENTIFICATION 4
CHAPTER 2. RESEARCH DESIGN 6
2.1 RESEARCH OBJECTIVE,RESEARCH QUESTION, AND SUB QUESTIONS 6
2.2 THROUGHPUT TIME 7
2.3 PRELIMINARY CAUSE AND EFFECT DIAGRAM 10
2.4 CONCEPTUAL MODEL 12
CHAPTER 3. METHODOLOGY 15
3.1 RESEARCH TYPE 15
3.2 RESEARCH METHOD 16
CHAPTER 4. DIAGNOSIS 19
4.1 VALIDATION OF THE BUSINESS PROBLEM 19
4.2 PROCESS-ORIENTED ANALYSIS 24
4.3 EMPIRICAL AND THEORETICAL ANALYSIS 31
VI
CHAPTER 7. FUTURE RESEARCH 58
REFERENCES 59
APPENDIXES 63
APPENDIX I:DESCRIPTION OF THE FUNCTIONAL DISCIPLINES 63
APPENDIX II:MANUFACTURING THROUGHPUT TIME PER PART REDUCTION 64
APPENDIX III:STEPS TAKEN TO MODIFY RAW DATA 65
APPENDIX IV:THE CURRENT VALUES STREAMS 68
APPENDIX V:DATA OF THE WORK IN PROCESS MEASUREMENT 83
APPENDIX VI: TRENDS IN QUEUE SIZES AT DIFFERENT WORKSTATIONS 84
APPENDIX VII:ANALYZED VALUE STREAM (OFFICIAL PROCEDURE) 85
APPENDIX VIII:BOTTLENECKS PRESENT IN THE CURRENT BUSINESS PROCESS 94
APPENDIX IX:DEVIATION IN DISTRIBUTION THROUGHPUT TIMES 96
APPENDIX X:REDESIGNED ADMINISTRATIVE BUSINESS PROCESS IS&SDELFZIJL 97
APPENDIX XI:ORDER INTAKE CHECKLIST 98
VII
LIST WITH ABBREVIATIONS
Abbreviation AAD BPC CPD CRM EBP ERP HRM FICO InEx IS&S JIT OSM SLA S.O. STS TECO TOC TQM VSM WACC WIP Meaning
Actor Activity Diagramming
Business Process Change
Chemical Park Delfzijl
Customer Relationship Management
Electronic Bill Payment
Enterprise Resource Planning
Financial Completed
Human Resource Management
In Execution
Integrity Solutions & Services
Just in Time
On Site Machining
Service Level Agreements
Service Order
Stork Technical Services
Technical Completed
Theory of Constraints
Total Quality Management
Value Stream Mapping
Weighted Average Cost of Capital
1
Chapter 1.
INTRODUCTION
This chapter is an introduction of the company Stork Technical Services (STS) located at Delfzijl, and helps to gain a better understanding of this research. A short introduction is given of the organisation Stork B.V., together with its history. A brief description is given about the core business is of Stork B.V., and the services that are provided. The different services that can be provided by STS Delfzijl are briefly captured and the services on which this research is focused are described in more detail. Further, the high-over business process of STS Delfzijl will be shortly outlined and finally the context of the initial business problem will be presented.
1.1 Stork B.V.
Stork B.V. is a manufacturing and service providing organisation that was founded in 1868 by Charles Theodoor Stork. The initial market on which Stork B.V. focused was the textile production. In 1954 Stork merged with “Werkspoor” (settled in Amsterdam) and went further under the name “Verenigde Machinefabrieken”. In 1968, after a mass dismissal the name changed to ‘’Verenigde
Machinefabrieken Stork”. In 1992 the company name changed back to Stork B.V., and in 1996 Stork
established an aero- and space department through the acquisition of Fokker Aviation. Today Stork B.V. is a multi-national organisation which consists of nearly 50 companies represented all over the world. Together all those companies accommodate employment for approximately 13.000 employees.
Now days Stork B.V. consists of two main groups; Stork Aerospace’s and Stork Technical Services. Stork Aerospace is responsible for designing, developing and producing advanced structures and electrical systems for the aerospace and defence industry. Stork Aerospace also supplies integrated maintenance services and products to aircraft owners and operators. The Aerospace group performs these activities with about 3,500 employees. The Technical Services group is a knowledge intensive, professional supplier of integrated technical services for installations and machines in the industrial market. The focus lies on creating added value by optimising the total cost of ownership. The basic principle is ensuring maximum availability of installations/machines at transparent (lowest possible) costs per unit of product through the entire lifecycle of the installation/machine1. The technical services group represented in the Netherlands execute these activities with about 9,300 employees and had a turnover of approximately € 1 billion in 2010.
1.1.1 Stork Technical Services
Stork Technical Services (STS) is represented as a world class, knowledge-based asset integrity management company, which is dedicated to providing seamless performance to corporations and asset owners in the Chemical, Oil & Gas and Power industries as well as railway and food & pharmaceutics industry. This research was conducted at STS Delfzijl, which is established in the Northern region of the Netherlands, and mainly focuses on providing services to customers located in that region. STS Delfzijl is located at Chemical Park Delfzijl (CPD), where it provides its services to almost every chemical plant established at the park. The CPD is the largest chemical complex of North-Netherland, and contains nine large chemical plants. The chemical plants that are represented on the CPD are presented in Table 1-1 together with their main products.
1
2
Company Main product
1. AkzoNobel Salt
2. Lubrizol Chloride-PVC
3. Bio MCN Bio methanol
4. MEB Chloride from salt
5. Delesto Steam and Power (current)
6. Teijin Aramid Aramide polymers
7. MCA Mono-chloride-acetic-acid
8. Delamine Ethylene amines
9. Bruner Mond Soda
Table 1-1 Companies located on the CPD
Three core activities are provided by STS Delfzijl. The first core activity is the assembly, installation, (preventive) maintenance, repair, renovation, modification, relocation and extensions of all kinds of technical installations, ranging from process installations to process supporting installations and machines. The second activity is consultancy and technical (maintenance) management. The third core activity is the management and execution of projects.
1.2 High-over Business Process
At most departments of STS Delfzijl an order is processed more or less at the same procedure. To get a little familiar with the procedure of processing an order, a general picture is drawn of the main processing steps. Davenport (1993) defines a business process as follow: “A process is a simply
structured, measured set of activities designed to produce a specified output for a particular customer or market. A process is thus a specific ordering of work activities across time and place, with a beginning, an end, and clearly identified inputs and outputs” (p. 5).
Figure 1-1 High-over process
Figure 1-1 presents the general procedure how an order is processed at STS Delfzijl. A brief description is given of the order routing, from the moment a request is made by the customer till the invoice is send to the customer:
The process starts with a request of the customer to provide a particular service within a certain matter of time.
After the request is done, there is determined whether the desired service can be provided by STS Delfzijl. When the responsible department is able to deliver the requested service(s), the order is accepted.
Now the administrative preparation of the order can be performed. The administrative preparation of an order exists of composing the work-instructions for the mechanic. Also, the components needed for delivering the requested service are ordered, after which the order is scheduled. Sometimes additional activities have to be performed during the preparation of an order, such as hiring a sub-foreman or arranging scaffolding.
Then a mechanic executes the physical work of an order to provide the requested service.
3 service. When performance or safety tests are conducted, the corresponding results are added in the report. Hereinafter, the customer has to sign a technical completion certificate to agree with the executed activities, which were needed to provide the requested service.
When the customer has agreed and signed the technical approval form, the order can be administratively finished by composing a cost specification. This summation of costs is send to the customer who has to approve it, before the actual invoice can be send.
The final output of the business process at Delfzijl exists of three elements: (1) the delivered service, (2) a technical report that lists the activities that are performed during the execution of the service, together with the technical specifications, and when necessary test results, (3) the invoice. The order is fully completed after the customer has paid the invoice.
1.3 Business Lines
STS is an integrated company, which is organized and grouped along core competences to offer full dedication towards the customers goals, and flexibility to meet the needs of the customers. The core activities are delivered by a range of different disciplines, divided in several business lines. In total nine business lines are defined in the business model (Figure 1-2). The model consist of seven business lines offering global leadership2: Asset Management, Consultancy, Integrity Solutions & Services, Process Equipment, Rotating Equipment, Electric Equipment and Sales & Rental Services. The other two business lines, Mechanical Services and Electrical and Instrumentation Services, are led on a regional base. Each of the individual business lines contains several product lines, which represent the services that are provided by the corresponding business line.
Asset Inegrity
- Project Management Services - Maintenance Management Delivery - Turnaround Management - Relocations - Brownfield Engineering - Strategy & Processes - Reliability & Integrity - Asset Management improvement
Business Lines Product Lines
Asset Management
Consultancy
- Turbo Machinery - Turbo Blading - GearBox Repair Services - Drive Trains & Auxiliary Equipment - Rotating Equipment Diagnostics - Electric Machinery Services - High Voltage Services & Diagnostics - Engineering & Maintenance Support - Boiler Services
- Deaerators & Swirlflash® - Burners & Retrofits - Prefab Piping & Skids - High Pressure Equipment - Process EPQ Projects - Energy Efficiency Solutions - Pumps & Valves Services - On-Site Machining & Bolting Services - Manufacturing Services - Inspection & Testing - CooperHeat Equipment - CooperHeat Services - Measurement & Control - Research Solutions
- Calibration and Validation Services - Industrial Equipment Rental Services - Industrial Equipment Sales - Industrial Equipment Education - Automation Services - Instrumentation - Electrical Installations - Electrical Building Systems - Calibration and Validation Services - High Voltage Service & Diagnostics
Solutions
Rotating Equipment
Electrical Equipment
Process Equipment
Integrity Solutions & Services
Sales & Rental Services Mechanical Services
Electrical and Instrumentation Services
Figure 1-2 Business model Stork Technical Services
2
4 1.3.1 Integrity Solutions and Services Delfzijl
The business line this research focuses on is Integrity Solutions & Services (IS&S) Delfzijl. Most services provided by IS&S Delfzijl are related to the repairing of products containing rotating parts. Rotating equipment is the general classification of mechanical equipment that is used to add kinetic energy to a process. The rotating equipment may be needed to move liquids, solids or gasses through a system of drivers (turbines, motors or engines), driven components, (compressors or pumps), transmission devices (gears, clutches or couplings) and additional equipment (lube and seal systems, cooling systems, buffer gas systems). The product lines represented by IS&S Delfzijl are pumps, valves and on site machining & bolting services:
Pumps: This product line includes work dealing with the repairing of centrifugal and positive displacement pumps, as well as other site equipment containing rotating components, such as blowers and extruders.
Valves: This product line includes work dealing with the process of valve maintenance, from dismantling, overhaul and repair right through to reinstallation, commissioning and. The replacement of control, shut-off and safety valves.
OSM & Bolting Services: This product line offers any imaginable machining operation that can be carried out on-site using mobile equipment.
Another part of the IS&S Delfzijl is the tensioning department. Here, components and materials are customized so they can be used by the mechanics to repair broken objects. Several departments are involved with the execution of the business process of IS&S Delfzijl. Appendix I gives a short description of the functional disciplines that are somehow engaged in the business process.
1.4 Problem Identification
The management of STS Delfzijl is unsatisfied about the current administrative procedure of processing an order at the IS&S business line. In the current situation it is not exactly known how the administration of an order is processed. There are many ambiguities about the way an order is administratively processed before and after the requested service is delivered. There is chosen to conduct the research at the IS&S business line because this administrative business process is representative for most of the business lines within STS Delfzijl. Furthermore, management believes that the deviation in time between the deliverance of the service and sending the invoice to the customer is the largest at IS&S Delfzijl.
1.4.1 Problem Definition
During two introductory meetings with management and some employees a number of issues were appointed about which they were dissatisfied. According to them the efficiency of the administrative process at IS&S Delfzijl is not optimal. The management has a strong presumption that the administrative business process is not performed in an uniform manner. It is to be expected that this results in several time-related problems:
Long and uncertain throughput time of the administrative business process.
Bad due date performance of the invoice (and technical documentation).
5 Three different stakeholder groups can be identified for this research, all having different requirements regarding the solutions of the mentioned problems. By appointing the requirements of the various stakeholder groups, clear expectations can be formulated (Table 1-2).
Stakeholder (groups) Requirements
Customer, maintenance operation managers, and sales
Faster order completion
No misfit service request and service delivery
Zero faults invoice and technical reports
Process owners (planners and foremen) Lesser administrative redundancies
Region and operational management Faster order completion
In control and manageable business process
Lesser (interest) costs
Table 1-2 Identification of the stakeholder requirements
The customers, maintenance manager, sales personnel and operation management would like to see that the order is faster completed, so the order can be financially closed. They complain about the large amount of time it takes before an invoice is send to the customer after the order is technical completed. Also the number of misfits between the delivered service by IS&S Delfzijl and the requested service by the customer should be reduced. A deviation between the requested and delivered service may result in additional throughput time. Also lesser faults in invoices and technical documentation are desirable. Errors made in the invoices or technical reports may result in rework and a longer throughput time. Planners and foremen complain about a the large administrative burden during the processing of an order. A lot of valuable time is lost because of administrative redundancies. According the management it is very hard to control and manage the status from orders (cost and time) of the business process. They would like have a more controllable and manageable business process. Management also would like to see a faster financial completion of the order. They would like to see that whenever a service is provided the accompanying technical documentation and invoice can directly be send to the customer. The management has a strong feeling that a large difference in time exist between the technical and administrative completion of an order. Finally, the management wants to reduce the costs that are made by administrative burdens and interest losses.
6
Chapter 2.
RESEARCH DESIGN
The second chapter begins with a formulation of the research object, research question and sub questions. Subsequently a description is given of the preliminary cause and effect diagram. Finally, theoretical perspectives are appointed which gives an indication of the scope and focus of this research.
2.1 Research Objective, Research Question, and Sub Questions
The two intake meetings revealed that management was dissatisfied about the efficiency of the current business process, which was caused by wasting precious time on the performance of non-value adding activities. This results in an throughput time of the administrative business process that is (too) long and uncertain. When appointed in terms of a functional problems, it results among others in dissatisfied internal and external customers due to a bad due date performance of the invoice (and technical documentation). In consultation with management the following business problem was selected; the throughput time of the administrative business process is too long.
2.1.1 Research Objective
Derived from the selected business problem the following research objective is formulated:
“Decrease the average throughput time of the administrative business process at Integrity Solutions & Services Delfzijl”
Decreasing the throughput time will improve the efficiency of the administrative business process when it is expressed only in time units. When a business process is more efficient, one uses a smaller amount time to execute the same activities. Efficiency does not only include the time dimension, but it is a wider concept. It is intended to attain the research objective without the need to invest in disproportional money and resources which will eventually have a negative influence on the efficiency.
2.1.2 Research Question
Given the selected research objective and the business problem the following research question is formulated:
“How can the throughput time of the administrative business process at Integrity Solutions & Services
Delfzijl be decreased?”
Because the research question is too broad and complex, its answer has to be found in several steps. In order to give an answer to this research question, some more narrower focused sub questions are used. These sub-questions emerged during the course of the research.
Sub question 1
Among the stakeholders there is no clear picture present of the current business process. It is not always clear what exact activities must be performed, necessary to process an order. To identify the main actions making it impossible to obtain a satisfying throughput time, the value stream of the entire business process should be mapped. This will give a good overview from the actual administrative business process. The first sub question is formulated as follow:
7
Sub question 2
The throughput time consist of the four components; process, setup, move, and waiting time. The setup, move and process times of the various administrative tasks are not known. To determine which of the four components have a great impact on the throughput time of the administrative business process the influences of the components should be measured.
“Which components have a great effect on the throughput time of the IS&S business line?”
Sub question 3
It became clear that the sum of process, and move times are only a fraction of the total throughput time. Excessive differences between processing time and throughput time are often caused by high work backlogs or by people creating artificial bottlenecks (Wiegand and Franck, 2005). By measuring the sizes of queues at the all workstations a good impression is created of locations where backlogs and bottlenecks arise, and waiting time is created.
“Which workstations from the administrative business process can be identified as a bottleneck?”
Sub question 4
All wasteful activities performed during the administrative business process must be identified. It is essential to analyse the current business process carefully, to identify causes of the long throughput time. To find all the rocks and stones that obstruct the order flow a detailed process analysis need to be conducted.
“Which wasteful activities of the administrative business process can be recognized during the processing of an order?”
Sub question 5
The fifth sub question focuses on the causes of the long throughput time, that are experienced by stakeholders which participate in the order processing. These variables can be very useful sources for improvement since the participants are the most acquainted with the business process (Nicholas, 1998). Sub question five aims to obtain as much as possible operational facts, so the validity of encountered problems can be claimed. Also the relations between these causes and the throughput time should be clarified by answering the fifth sub question.
“What relevant causes of (and their relationships to) the long throughput time can be identified?”
Sub question 6
In the diagnoses chapter it is explained that the bad order intake, unclear business process and the usage of two independent IT systems are the most significant problems. Also a lack of discipline and sense of responsibility seem to cause a long throughput time. By answering sub question six the most proper interventions are found to tackle the causes, and improve the throughput time.
“Which interventions should be taken in order to design an improved future state map?”
2.2
Throughput Time
8 underappreciated in many organizations. Optimizing it will lead in many cases to improvements that are just waiting to be discovered, and considerable costs can be saved (Wiegand and Nutz, 2008). A production process is regarded as a transformation process, because raw materials are converted into finished products by operators using machines and tools. An administrative process can also be regarded as a transformation process, in which data is processed by people with the aid of information systems (Bertrand, Wortmann and Wijngaard, 1998). While in a factory a material product is produced from tangible raw materials, in administration an input (the information, such as customer data or purchase orders) is also turned into an output (the service, a processed order, a quote) via a production process. (Wiegand and Nutz, 2008).
There are a lot of complaints about the long throughput time of the administrative business process. The throughput time is the average time from the release of an order at the beginning of an routing until it reaches an inventory point at the end of the routing (Hopp and Spearman, 2001). It is a random variable that gives the time an individual job takes to traverse a certain routing (Hopp and Spearman, 2001). The administrative business process of IS&S Delfzijl consists of two separate stages, a preparation stage and a completion stage. The preparation and completion stages are separated by the physical execution of the order (providing the requested service). Despite the assumption that the throughput time of the completion stage of the administrative business process is too long, the entire business process will be analysed. Activities performed during the preparation stage of the business process may affect activities performed during the completion stage, which eventually affect the total throughput time. The total throughput time of the business process is measured from the moment the order is accepted till the invoice is send to the customer.
2.2.1 Components of Throughput Time
Reduction of the throughput time can generate multiple benefits, such as lower work in process (WIP), improved quality, more satisfied customers, and less loss of interest. These benefits can be important factors for the profitability of STS Delfzijl. Reducing the throughput time of a process can be a daunting and confusing task due to the large number of factors that can be changed and the interactions between them (Johnson, 2003). To reduce the throughput time the basic factors that determine the administrative throughput time must be clearly understood. The framework of Johnson (2003) shows the factors that influence the manufacturing throughput time (Appendix II). Although this framework is intended to reduce the throughput time of an production process, it can also be useful to reduce the throughput time of an administrative business process. The manufacturing throughput time is composed of components that also can be used to determine the administrative throughput time. The administrative throughput time is composed of four components; process, setup, move, and waiting time (Figure 2-1).
Figure 2-1 Administrative throughput time per order based on the framework of Johnson (2003) Process time: The process time is the sum of times needed to process the contents of all
9
Setup time: The setup time is the sum of times needed to obtain all information which is necessary to perform all work steps.
Move time: The move time is the time spent moving documents and/or information – needed to perform all the administrative activities during the business process – between working steps.
Waiting time: The waiting time is the sum of time an order is in a queue waiting to be processed, and the sum of time an order is waiting on information necessary for the work steps to be carried out.
2.2.2 Factors Affecting the Components
Under the most ideal conditions a person is not disturbed during the execution of his task, and all necessary information is available. However, there are often moments that an employee is disturbed from his administrative task, such as disturbed by phone calls, e-mails or passing customers. Also, when a colleague has a request for information (needed to perform his work), it results in idle time. Rework – which occurs every time a person has to go back to an earlier process step in de workflow – influences the throughput time of the administrative business process (Wiegand and Nutz, 2008). Several factors that could affect the components of the throughput time are shown in the third column of Appendix II. The factors affecting the components are converted from an manufacturing setting to an administrative setting, which are explained in Table 2-1.
Factors affecting the components
Brief description of factors for an administrative environment
Setup time The time necessary to obtain all task-related documentation and/or information.
Number of setups The number of steps necessary to obtain all task related documentation and/or
information.
Scrap and/or rework The adjustments or renewal of outputs of the administrative business process
(such as; invoices, quotations, work instructions, etcetera).
Number of operations The number of tasks needed to process the entire administration of the order.
Operation time The total time required to execute all administrative activities of an order.
Moving time The total time required to transfer documents and/or information between all
workstations.
Number of moves The number of moves required to transfer documents and/or information
between all workstations.
Batch size The quantity of similar jobs that can be administrative processed in a series (for
example; a series of invoices that is processed sequentially).
Transfer batch side The quantity of similar jobs that are transferred from one workstation to the
next workstation.
Processing variability
Variability in processing time comes from several sources: variance in setup time for a workstation, variance in the processing time per order, and variance in the size of the batch processed.
Arrival Variability
Is dependent on the variability of new orders released directly at the workstation, as well as the departure variability from any upstream workstation that feed the station in question.
Workstation utilization
The total workstation time required per period divided by the total workstation time available per period.
Resource access
The availability of resources which determines whether or not there has to be waited during the administrative processing of an order. This could include hiring workers, overtime, access to all documentation and/or information, etcetera.
Number of queues Depends on the number of orders at a workstation waiting to be processed
10
2.3
Preliminary cause and effect diagram
Several interviews were conducted to gain a better understanding of possible actions that may alter the factors of the administrative throughput time in the situation of the IS&S Delfzijl. People representing different disciplines who are somehow affected by the business problem, were interviewed (Corbin and Straus, 2008). The information gained during the interviews is structured and presented on a conveniently way in the form of a cause and effect diagram (van Aken, Berends and van der Bij, 2009). The preliminary cause and effect diagram (Figure 2-2) identifies potential causes of the long throughput time in the administrative environment of IS&S Delfzijl. The preliminary cause and effect diagram is partially derived from the manufacturing throughput time per part reduction framework of Johnson (2003), but is related to the selected business problem. The relations shown between the potential causes and the factors listed in column 2 are determined before the actual research is performed and are of conceptual nature. The potential causes mentioned in the interviews do not necessarily have to exist in the current administrative business process of IS&S Delfzijl. The preliminary cause and effect diagram is only based upon interviews, so there is no guarantee that the problem and its potential causes are valid and reliable. The diagram is only intended to give an indication of some directions that will be explored (van Aken, Berends and van der Bij, 2009). The actual causes of the long throughput time will be explored and validated during the problem diagnoses stage (van Aken, Berends, and van der Bij, 2009).
11 In the second and third column of the preliminary cause and effect diagram the factors and components that affect the administrative throughput time are depicted. The potential causes listed in the first column are actions that could affect one or more factors of the administrative throughput time. During the meetings a number of dissatisfying actions (perceived in the current administrative business process) were appointed that may influence the throughput time. With the guidance of reliable literature these potential causes of a long throughput time are assigned to one or more factors listed in the second column of the preliminary cause and effect diagram. Due to the many possible relationships between the causes and factors the model will soon be indistinct by the many connecting lines. To keep the model clear and understandable all possible causes are circled and refer to all factors. The actual relations between the causes (present at IS&S Delfzijl) and factors are explored and validated during the problem diagnosis stage.
In an administrative setting, individual employees carry out a large variety of different tasks, and are involved in numerous service delivery processes with varying levels of complexity requiring various degrees of expertise (Wiegand and Nutz, 2008). During the meetings it was mentioned several times that the current procedure of processing an order is complex and not transparent. In administrative areas design and performance of work is often left to the individual, and a person gives his own interpretation (as he thinks is best) to the execution of tasks (Wiegand and Nutz, 2008). This all may end in “typical” problems found in indirect areas in the company and service delivery processes, queue and quality problems, rework waiting times and idle times (Wiegand and Nutz, 2008). An unclear procedure of processing an order can quickly lead to a high amount of operations, moves, queues, and setups. An complex procedure of processing orders can also mean that the routing of different orders can greatly vary. A so called job shop is characterized by strong fluctuations in demand for capacity per operation type or workstation (Bertrand, Wortmann, and Wijngaard, 1998). In a job shop there are constantly changing production bottlenecks where queues may arise, whereby the variation in arrivals is an important factor (Bertrand, Wortmann, and Wijngaard, 1998). Working orders can have strongly varying occupation times, and thus affect the processing variability. The time taken to perform the administrative tasks varies considerably and ranges from simple tasks that can be performed in 10 seconds to resolving a complex problem which takes several hours to complete (Wiegand and Nutz, 2008).
A complexity of services is heterogeneity whereby there is a high chance of idiosyncratic customer expectations and perceptions (Prajogo, 2006). When the delivered quality does not match the expectations of the customer, it causes rework or scrap. By causing work to be done over (either as rework or as replacements for scrap), quality problems often end up in increasing the utilization of workstations. (Hopp and Spearman, 2001). By affecting both variability and capacity, quality problems can have extreme operational consequences (Hopp and Spearman, 2001). Also when quality problems causing work to be done over the total setup and process time of an order will increase, and new queues will arise.
12 several administrative activities. The absence of an all-embracing ERP system (to perform all administrative tasks can be performed) might increase the operating and setup time of an order, because the possibility of double work. Also the number of operations, setups and queues might be larger when activities are executed using of multiple ERP systems. Not everyone who is involved in the administrative business process has access to all documentation/information, because one of the ERP systems is a stand-alone system that is only used at the IS&S business line.
The objective of managing raw materials (in this case raw data) is to have them available when needed by the production process (respectively the administrative process) without carrying any more inventory than possible (Hopp and Spearman, 2001). In the case of services goods are produced, sold, and consumed, commonly occur at the same place and time (Prajogo, 2006). Buffers cannot be prepared between production and consumption stage. When information needed to perform administrative tasks (in support of services) is not directly available there must be searched for it, which increases the setup time of an order. When the information is wrong or incomplete mistakes are easily made, increasing the amount of rework and scrap. When an person has no authority to perform certain administrative activities, it will affect the availability of resources. The presence of customers brings complexity to management of service operations. Firms do not only have to pay attention to “end services”, but also to the process of rendering and delivering those services (Prajogo, 2006). A lack of communication with the customers may result in a low perceived service quality, causing rework or scrap. When the customer is not involved during the execution of the service, and findings are not fed back unnecessary operations, moves and setups may be performed.
A high level of WIP results in rework and scrap, which can be illustrated with the river of inventory analogy (Bertrand, Wortmann, and Wijngaard, 1998). In this analogy the business process is conceived as a river, where an uninterrupted production means that a boat can sail unflustered down the water. There is no visibility of damaging rocks, which disguises knowledge of underlying dangers. When WIP level at the work-floor is high, problems can also be hidden in a similar way. By releasing new orders on the work floor and incorporating time buffers, potential production (or administrative) problems will not be discovered, causing more rework and scrap. With a high level of WIP the utilization of the workstations will also become larger. The utilization of a workstation may strongly depend on the priority of orders. When a certain workstation has a low priority, the amount of orders waiting to be processed can increase rapidly. The throughput time is mainly determined by waiting on available capacity at the different workplaces (Bertrand, Wortmann and Wijngaard, 1998). Due to the lack of capacity choices must be made, which activities have the highest priority. When order have to wait at the availability of capacity to be processed, the number of queues and the utilization of workstations will become larger.
2.4 Conceptual model
13 theoretical perspectives that are listed in the conceptual model are used to study the business problem and are likely to be important during the diagnosis and/or design phase.
Figure 2-3 Conceptual model based on the black box model of de Leeuw (2002)
Six theories or policies were selected using reliable literature, so the detrimental causes can be managed:
Process Innovation and Quality oriented theories
Knowledge and Information Management theories
IT policies
Human Resource Management related policies
Production Planning and Control
Customer Relation Management related policies
According to Davenport (1993) the term process innovation encompasses the envisioning of new
work strategies, the actual process design activity, and the implementation of the change in all its complex technological, human and organization dimensions (p. 2). It is a structured approach that
accompanies radical changes in relation to the current business process. Process innovation is a discrete initiative that must be combined with other initiatives for on-going change (Davenport, 1993). Quality oriented methods can be one of those initiatives, in order to avoid that a company returns to old ways of doing business. Quality based improvement is a valid way to achieve incremental process improvements on a broad scale (Davenport, 1993). Quality oriented methods are step-by-step approaches of streamlining the business process. A well-known example of a quality based policies is total quality management (TQM). The essence of TQM is that major improvements in the process effectiveness can result from a continuous series of incremental efforts, directed towards identifying and removing all sources of process (Hayes, Pisano, Upton and Wheelwright, 2005). Although both procedures have much in common (both focus on improvement of the process and emphasize the importance of top management support), there are also some important differences (Hayes, Pisano, Upton and Wheelwright, 2005). For process innovation it is assumed that the process being reengineered is understood well enough (learning before doing) to be completely redesigned while quality based procedures typically employs learning by doing to encourage improvements to existing processes throughout the organisation (Hayes, Pisano, Upton and Wheelwright, 2005).
14 2004). It is the systematic and active managing of ideas, information, and knowledge redesign in an organization´s employees. The structuring of knowledge enables effective and efficient problem-solving, dynamic learning, strategic planning, and decision making (Turban, Aronson and Liang, 2004). Information management is defined as the overall management of a company’s entire information environment (Davenport, 1993). There should be some process to warrant the usage of a proper information system during the administrative order processing. An information system collects, processes, stores, analyses and, disseminates information for a specific purpose (Turban, Aronson and Liang, 2004). Information system policies are dependent of an interaction between technology and other aspects (people, structures and processes) within a company. These policies should express which IT system should be used during each task, and can be influenced by the quality improvements.
Human resource management (HRM) refers to the policies, practices, and systems that influence employees’ behaviour, attitudes, and performance (Noe, Hollenbeck, Gerhart and Wright, 2008). Several important HRM policies can affect the company’s performance. Special attention is given to the training policy, since this is an important policy where quality oriented methods can be implemented. Quality oriented methods enable a culture of continuous improvement, also known as kaizen. By gaining experience employees can experiment with different ways of doing it to see if their performance improves (Hayes, Pisano, Upton and Wheelwright, 2005). Kaoru Ishikawa says: “Quality
control begins with education and ends with education” (Nicholas, 1998). A decent training should
help to create a culture where employees are aware of continuous improvements. It is assumed that other important practices that can be considered in relation with human resources (such as: design of work, recruiting, performance management, etcetera) are covered by HRM.
15
Chapter 3.
METHODOLOGY
In order to conduct a feasible and appropriate study, the general context of the research must be indicated. This chapter starts with a classification of the type and conditions of this research. Succeeded by a description of the used method, which presents a more practical approach how to conduct this research operationally.
3.1
Research Type
The area of research on improving the throughput time of an administrative business processes within a service-oriented organization is an interesting subject for both basic theory and applied research. The problem of this research can be identified as a real life problem, for which it is possible to pursue the research on a theoretical level, but commonly practical benefits flow from it (Walliman, 2001). The value of this research will be measured on scientific contribution and practical applications. This research serves as a basis for further applied research or a design-focused business-solving project. (van Aken, Berends, and van der Bij, 2009). Design-focused problem-solving projects are practical projects in which the performance of a company, department or business system on several conditions is improved by the design of a solution (van Aken, Berends, and van der Bij, 2009).
In this research the instrumental perspective and hard system approach are assumed. With the instrumental perspective the management is considered as a purposeful control of processes, and organisations as instruments to achieve objectives (de Leeuw, 2002). The hard system approach is based on a clear problem situation where it is relatively easy to find an unequivocal answer (de Leeuw, 2002). The researcher regards the problem situation as independent and predictable. However, when the problem situation is examined the background of the researcher is influential. Practical problems should be viewed from more perspectives to prevent unilateral spurious solutions. (de Leeuw, 2002). A pluriform research will improve the objectivity, because multiple respondents with different (functional) backgrounds are more reliable than a similar number of respondents with the same (functional) background. Both the persons interviewed as the persons who participated in the workshops have different functional backgrounds and responsibilities. The functional backgrounds of the interviewees were planners, foremen, maintenance managers operation managers, sales managers, quality managers, representative of the financial and calculation department. During several workshop to map the value stream, analyse it and design a proper solution, representatives from different departments were present. The workshops were attended by planners, foremen, maintenance managers, representative from the F.E.B., sales managers, mechanics and operation managers.
16 for this research. Different from the assumption underlying the analytical approach, is that reality is arranged in such a way that the whole is not summative, so not only the parts but also their relations are essential (Arbnor and Bjerke, 1997). In this research a system is defined as a structured set of
objects and/or attributes together with the relationships between them (Wilson, 1990; p. 24). The
scientific ideal is to reach better explanations and understandings of how various types of systems behave under different circumstances (Arbnor and Bjerke, 1997). All parts and the relations between them need to be found, and every picture of the new system should be than better the last one (Arbnor and Bjerke, 1997).
3.2
Research Method
After the research type is specified a proper research method should be determined, how to achieve a good solution. The way how to make a decision methodological is called procedural rationality (de Leeuw, 2002). In order to decrease the long throughput time, it is expected that the solution of this research is a intervention or a redesign of a certain (part of a) business process. To solve this kind of business problems a methodology with a designed oriented nature should be used. The Diagnose, Design, Change (DDC) method is a systematic procedure to address management problems (de Leeuw, 2002). Similar to this method is the Design Focused and Theory Based Business Problem Solving (DFTB BPS) methodology of van Aken, Berends and van der Bij (2009). The DFTB BPS method is following the logic of a problem solving (regulative) cycle, consisting of six steps: (1) the problem mess, (2) the problem definition, (3) the analysis and diagnosis, (4) the formulation of the action plan, (5) the intervention, and (6) the evaluation. The evaluation step subsequently leads to a new problem mess, from which the cycle is deployed again. In this study the last two steps of the problem-solving cycle are excluded because a lack of time. The remaining four steps correspond to the steps taken in the DDC model. In the context of this research the DDC model of de Leeuw (2002) is used to characterize the procedural rationality. In both cases a pluriform diagnosis and a comprehensive problem diagnosis is emphasized.
3.2.1 Diagnose, Design, Change Model
The preliminary cause and effect diagram and selected business problem, shown at the top of the DDC model (Figure 3-1) are input of the problem diagnosis. The left side of the model shows the different steps that are taken during the research, and the right side of the model shows the tools, literature, and concepts used to execute all different stages of the research. The right causes for a long throughput time are selected with the aid of the naming and framing process. These causes will be subject of the interventions that need to be taken to solve the business problem.
Diagnosis
17 Figure 3-1 DDC Model
The process oriented analysis is used to investigate the actions (causes) that influence components of the throughput time (the problem variables). The practical way of executing all operation of the business process is carefully mapped. During multiple workshops the current value stream of IS&S Delfzijl is mapped, with the help of people from the work floor. Every participant of the workshop is somehow involved with the execution of orders related to business process of IS&S Delfzijl. The people who are daily involved with the business process have a good understanding of the actions that need to be performed to completely process an order. The modeling tool that was used to map the current value stream is Actor Activity Diagraming (AAD). AAD is modeling tool (Schaap, 2001) in which a business process is decomposed in terms of transformations and transactions. Each single activity performed during the processing of an order has been mapped. For each of the activities is determined which actor brings the product or service from one state into the next state perform the transformations. The transactions are transitions from the product or service in a defined state from one actor to another actor (Schaap, 2001).
18 and calls them queries, rework loops and fields of action (“process bursts”) (Wiegand and Nutz, 2008). During several workshops the value stream is step by step analyzed to identify and indicate inefficient activities. Again this analysis is performed with persons representing different functional departments.
To obtain information and opinions from experts who are involved with the business problem several interviews are conducted (Walliman, 2001). Semi-structured interviews are used to keep the explorative nature of the diagnosis and limit the dependency of the researcher (van Aken, Berends, and van der Bij, 2009). A number specific questions regarding the problem statement are designed, while sufficient room is left for further development of the answers and additional information. Because the system approach is used are subjective interpretations in this research treated as objective (Arbnor and Bjeke, 1997). Information is obtained from that part of the population of which is thought they know the most about the subject (Walliman, 2001). From the pluriform perspective several persons with different views were selected to be interviewed. The people that were interviewed represented all functional departments that somehow are involved or linked with the (administrative) business process of IS&S Delfzijl. All functional departments are at least covered by one respondent. The people involved with the business process are best capable to identify critical issues regarding the business problem.
Also several other research instruments are applied, for the purpose of triangulation (van Aken, Berends, and van der Bij, 2009). The researcher was on a daily basis present at the office of IS&S Delfzijl. The concept observation is used here in its philosophical sense, referring to the use of human senses to produce evidence about the ‘empirical’ world (Blaikie, 2003). Participative observation enables the researcher to experience organizational processes from within (van Aken, Berends, and van der Bij, 2009). There are many formal (workshops or weekly meetings) and informal (lunch or coffee breaks) moments where extensive and relevant discussions take place. Also many secondary data will be collected and analyzed; such as answers of former interviews, quality related sites (KAM) at the intranet, planning manuals, contracts (Service Level Agreements) and other documents, and information systems. Extensive measurements were performed to monitor the course of all orders through the business process, to identify causes of the long throughput time.
The possible causes of the long throughput time, that were experienced by members of the business process, are further explored in the empirical analysis. By means of the empirical analysis causes of the long throughput time are identified, and evidence is gathered to support the diagnosis. In the theoretical analysis existing literature is used as an extra source of evidence of the causal relationships. By using reliable literature lines of causes are validated (van Aken, Berends, and van der Bij, 2009). As well the preliminary cause and effect diagram and the actual administrative business process will play an important part in the exploration and validation of the causes.
Design
19
Chapter 4.
DIAGNOSIS
To come up with the right solution to the business problem, it is important to perform an intensive diagnosis. A proper diagnosis should start with the validation of the selected business problem, and a fully-fledged, integral explanation of the business problem (van Aken, Berends and van der Bij, 2009). After the business problem is validated the process oriented analysis is started. First, the value stream of the entire business process of IS&S Delfzijl is mapped. Subsequently, the processing times and queue sizes are measured from all workstations involved in the administrative business process. Followed by a detailed analysis of the value stream on wasteful activities and other inefficient actions. Finally, the causes of the long throughput time experienced by the different stakeholders are examined in the remainder of the diagnosis. The causes are checked on validity and their connection to the business problem are investigated. Although the different analysis steps are described sequentially, the diagnosis is performed iteratively.
4.1
Validation of the Business Problem
To determine whether the average throughput time of the administrative business process is too long, the business problem is checked for validity. By comparing the actual throughput time of the administrative business process with norms there is determined whether the performance level is satisfactory or not. The validation of the business problem is used to further specify the business problem, so that the causes of the business problem can be traced more effectively (van Aken, Berends and van der Bij, 2003).
4.1.1 Norms
STS Delfzijl has concluded maintenance contracts with a number of customers located at the CPD. In a so called Service Level Agreement (SLA) all kinds of agreements are defined regarding maintenance work of factory related objects. Also agreements about the administrative settlement of orders are defined. In some of these SLA’s norms are included regarding the maximum number of working days it may take to send a cost specification to the customer after the requested service is delivered. A second norm agreed in the SLA includes the maximum number of working days within which a customer has to respond (reject or approve) at the cost specification after it is sent. In Table 4-1 the sizes of the norms are shown that are concluded with several companies.
Company Norm 1 (Time to
compose cost spec.)
Norm 2 (Time to respond cost spec.)
Akzo Nobel (Salt) 20 working days 10 working days
Delesto 20 working days 10 working days
MEB - 10 working days
MCA - 10 working days
Table 4-1 Norms concluded in SLA’s
Besides the comparison of the throughput time with the norms, it is also important to determine whether the measured throughput time is acceptable regarding the efficiency of the business process.
4.1.2 Modification Steps
20 planners and foremen to process all administrative tasks The data extracted from the Jerp administration system and SAP system is un-preprocessed, incomplete, and noisy. The raw data contains fields with missing values, and fields that are redundant or contain outliers. This raw data is modified, so useful information could be extracted. To conduct a reliable measurement representing the actual throughput time, several modification steps must be taken; such as data cleaning, handling missing data, and identifying misclassifications (Larose, 2005). In order to modify the raw data and present is a convenient way the following modification step are taken3:
1) Gathering and cleaning up the raw data. 2) Determine the right measurement points. 3) Deleting fields containing missing values. 4) Determine the throughput time per order. 5) Deleting fields containing outliers.
From the Jerp administration system raw data is gathered, existing of a list with service orders from all jobs executed in the year 2010 by IS&S Delfzijl. The data is cleaned up by deleting all service orders for which IS&S Delfzijl was not fully responsible and/or other business lines were involved. The next step was to determine the points between which the different throughput times (of the entire business process) could be measured. These measurement points are derived from the SAP system and are summarized in Table 4-2. A distinction is made between two types of service orders; order from customers making use of an Electronic Bill Payment (EBP) system, and orders from customers who do not make use of the system. For each individual service order (that remained after the last modification step) the throughput times are calculated between the different measurement points. By mapping the distribution of all throughput times between the various measurement points, outliers were identified and deleted. This resulted in 158 and 238 service orders from respectively non EBP using and EBP using customers, that could be used to measure the average throughput time.
Measurement point Description measurement point Measurement
point EBP
Measurement point non EBP Firstly book hour The date on which the first hour is booked on an
order. 1 1
Last booked direct hour
The date on which the last hour of a mechanic is
booked on an order. 2 2
Cost specification signed by customer
The date on which the cost specification is signed by
the customer. 3 3
Cost specification to financial department
The date on which the financial department receives
the signed cost specification from the customer. 4 4
Sending pro forma through EBP
The date on which a pro forma invoice is sent to the
customer via the EBP system. 5 -
Pro forma signed by customer on EBP
The date on which the customer agrees with the pro
forma invoice via the EBP system. 6 -
Invoice send to customer
The date on which the invoice is sent to the
customer. 7 5
Invoice paid by customer
The date on which the invoice is paid by the
customer. 8 6
Table 4-2 Measurement points of the throughput time
3 Appendix III describes in detail the modification steps that are taken to convert the raw data into information
21 4.1.2 Actual Throughput Time
After the raw data is transformed into useful information, the throughput times can be calculated. The average throughput times are calculated between the various measurement points, for orders from EBP using and non EBP using customers. The multiple throughput times are calculated using the following throughput time:
Equation (4-1): ( ) (∑ ( ) ) ( ⁄ ) The X in equation (4-1) represents the total number orders (238 for EBP and 138 for non EBP using customers) used to calculate the average throughput times. The n represents the number of the corresponding service order from the list with remaining orders after the modification steps. The i indicates the corresponding measurement point shown in Table 4-2, as well for EBP using customers as non EBP using customers. TT(i)n represents the throughput times that were calculated with
equation (III-1) in Appendix III. Finally, the days are converted to working days by multiplying the average throughput time by ⁄ . The implementation of the list with modified data into equation (4-1) resulted in the throughput times that are shown in Figure 4-1.
Figure 4-1 Actual throughput times at IS&S Delfzijl
Figure 4-1 shows two additional measurement points compared with the measurement points listed in Table 4-2. These additional points are included to visualize the norms that are agreed in SLA’s with some customers. The measurement point “last booked hour” is important to determine the time it takes to compose a cost specification after the requested service is delivered. This measurement point could not be identified out of the raw data, obtained from the SAP system4. The point “cost
specification to customer” could also not be obtained from the SAP system. This measurement point
is necessary to determine the number of working days it takes to approve the cost specification by the customer, after it is sent by IS&S Delfzijl.
4 Appendix IXAppendix IX: Deviation in Distribution Throughput Times describes why this measurement point
