Modern Flying Factories i byggindustrin: En beskrivning av konceptet och lärdomar för fortsatt utveckling

(1)

Modern Flying Factories in the Construction Industry

A Description of the Concept and Lessons for Further Development

Sanna Haukka Magda Lindqvist

2015

Master of Science in Engineering Technology Architecture

Luleå University of Technology

Department of Civil, Environmental and Natural Resources Engineering

(2)

i

ABSTRACT

Off-site production is pointed out as a solution to the construction industry’s lack of

efficiency. A way to produce off-site, without a high capital cost, can be found in temporary factories. It is the basis of a research project, which uses a concept of ‘Modern Flying

Factories’ (MFFs). The concept is today novel and lacks a defined strategic positioning.

However, temporary factories are not something unique and similar concepts can be found in Swedish field workshops and Flexible Field Factories for the construction industry and the concept of Mobile Manufacturing for the manufacturing industry. The purpose of this study is to increase the understanding and facilitate future strategic development by

describing the MFF concept and suggest lessons for further development. The study is based on a literature review and a case study through interviews and observations of the concept in its current state with two pilot projects. The ‘Glenfrome project’ was the first pilot project in which timber and straw bale panels were produced for a school extension. Although the application of the concept managed to shorten the construction period, the project was considered to have potential for improved efficiency. The second pilot project, the ‘Battersea project’, produces utility cupboards to apartments in a major real estate development. This project was during the time of the study in its start-up phase and several difficulties were observed. The majority of them could be traced to the design.

This study presents a description of the concept in seven propositions. It proposes that MFF is an off-site production strategy, which on projects applies a standardized process for start- up, operation and close down of a temporary factory. Its competiveness against stationary factories lies in allowing flexibility in location, time and production technique through low capital investment. The objective is to ensure time and quality as well save costs. The possibilities for achieving these are however restricted due to not having a continuous production process. Two alternative strategic directions have been identified for the concept in the future: continue with low investment costs to maintain flexibility or to automate the production to increase the efficiency. The lessons in this study are founded in the pilot projects, as well as what could be learned from theory of similar concepts and

manufacturing improvement methods. Many of the lessons shows a need for an increased control over the start-up phase and are mainly concerning: decision making, early decisions and involvement of expertise, increased control over the design and production process and increased factory friendliness.

(3)

ii

SAMMANFATTNING

Prefabricering pekas ut som en lösning på byggbranschens bristande effektivitet. Ett sätt att producera off-site, utan en hög kapitalkostnad, kan vara temporära fabriker. Detta är grunden för ett forskningsprojekt som använder ett koncept med ‘Modern Flying Factories’

(MFFs). Konceptet är idag i sin pilotfas och saknar en definierad strategisk positionering.

Temporära fabriker är dock inte något nytt och liknande koncept kan hittas i svenska fältverkstäder och Flexible Field Factories för byggnadsindustrin samt konceptet Mobile Manufacturing för tillverkningsindustrin. Syftet med denna studie är att öka förståelsen och underlätta framtida strategiska utveckling genom att beskriva MFF konceptet och föreslå lärdomar för framtida utveckling. Studien bygger på en litteraturstudie och en fallstudie genom intervjuer och observationer på konceptet i dess nuvarande stadie med två pilotprojekt. ‘Glenfrome projektet’ var det första pilotprojektet där paneler av limträ och halm producerades för en skolutbyggnad. Trots att applicering av konceptet lyckades förkorta byggtiden, ansågs projektet ha potential för ökad effektivitet. Det andra

pilotprojektet, ‘Battersea projektet’, producerar installationsskåp till lägenheter som en del av en stor fastighetsutveckling. Detta projekt var under tiden för studien i sin uppstartsfas och flera svårigheter observerades. Majoriteten av dem kunde spåras till projekteringen.

Denna studie presenterar en beskrivning av konceptet som sju propositioner. Det föreslås att MFF är en off-site produktionsstrategi, vilket på projekt, tillämpar en standardiserad process för uppstart, drift och nedläggning av en tillfällig fabrik. Dess konkurrenskraft mot

stationära fabriker ligger i att erbjuda flexibilitet i lokalisering, tid och teknik genom låg investeringskostnad. Målet är att säkerställa tid och kvalitet samt spara kostnader.

Möjligheterna till detta är dock begränsade i och med att produktionsprocessen inte är kontinuerlig. Två alternativa strategiska riktningar har identifierats för konceptet i

framtiden: att fortsätta med låga investeringskostnader för att bibehålla dess flexibilitet eller att automatisera produktionen för att öka effektiviteten. Lärdomarna i denna studie grundar sig i pilotprojekten samt vad som kan läras från teori kring liknande koncept och

förbättringsmetoder inom tillverkningsindustrin. Många av lärdomarna visar ett behov av en ökad kontroll över uppstartsfasen och berör främst: beslutsfattande, tidiga beslut och medverkan av expertis, ökad kontroll över utformningen och tillverkningsprocessen samt ökad fabriksvänlighet.

(4)

iii

FOREWORD AND ACKNOWLEDGEMENT

This thesis is the final part of our studies to get a Master in Architectural Engineering, conducted for the institution of Civil, Environmental and Natural Resources Engineering at Luleå University of Technology (LTU). It has given us a valuable opportunity to achieve a deeper knowledge within construction, which is the area we are passionate about. The opportunity to study how industrial construction can be applied to a traditional

construction market has been inspiring and instructive. We believe that this meeting will become increasingly important in the future to achieve a more efficient construction.

The study has been carried out through an Erasmus student exchange, with the University of Reading, UK, as the hosting university. We are very grateful to have had the opportunity to live in another country with the inspiring meetings and challenges it has implied. We would therefore like to thank the University of Reading for giving us this opportunity despite all the problems that arose along the way.

The first persons we would like to extend a special thanks to are our supervisors. Dr Chris Harty, for always taking the time for us and giving us essential possibilities for creating this thesis. Professor Lars Stehn, for your valuable guidance during our work with the thesis.

We also want to express our gratitude to Andrew Skinner for giving us the opportunity to collect data and for taking time to share your knowledge. Last, but not least, we want to thank Martin Neeson, Robert McCulloch, Craig White and Jason Loomes for participating in the study. Without your involvement, this study would not had been possible.

Luleå April 2015 Sanna Haukka and Magda Lindqvist

(5)

iv

EXPLANATION OF TERMS

This study handles several concepts of temporary or mobile production capacity. This section gives a short explanation of the concepts concerned and gives the abbreviations used further in this study.

Modern Flying A concept for the construction industry with temporary Factory (MFF) manufacturing in a factory environment. The concept is in a

developing phase and performing pilot projects.

Modern Flying Temporary factories used for production in projects Factories (MFFs) that applies the MFF concept.

Flying Factory (FF) A concept for production of straw bale panels in temporary factories. Has been the basis for development of the MFF concept.

Flexible Field A conceptual production concept for the construction

Factory (FFF) industry, which offers mobile and reconfigurable production in containers with fully automated production processes. Has not yet been applied in reality.

Mobile Manufacturing A concept for the manufacturing industry, which offers

(MM) mobile and reconfigurable manufacturing in containers with the possibility to vary the level of automation in the production process.

Swedish field Temporary factories on the construction site, which started factories (SWF) to be used during the 60s and function as a workshop. The

temporary factories are simple and could be tents or already completed buildings at the construction site such as garages.

(6)

v

1 INTRODUCTION

1.1 Background

For some time, the construction industry has been criticized for its low efficiency.

Reports such as “Rethinking Construction” (Egan, 1998) by Construction Task Force in the UK and “Skärpning gubbar!”(SOU 2002:115) by Byggkommissionen in Sweden has highlighted deficiencies such as low profitability, low quality and a lack of radical change. For improvements in the construction industry, Egan (1998) identified off-site production through pre-assembly and prefabrication to be of significance.

Over the years, many authors have written about off-site production and its contribution to the construction industry. The theory, and particularly around industrialization, has largely been influenced by methods from the manufacturing industry (e.g. Barlow et al., 2003; Martínez, Jardón, Victores & Balaguer, 2013; Jonsson

& Rudberg, 2014). Design for Manufacturing and Assembly (DFMA) is considered as an important tool for achieving a cost effective manufacture by evaluating and simplifying the product structure (Bogue, 2012). Lean philosophy is often seen as one of the main foundations of industrialised processes in the manufacturing, as well as the construction industry (e.g. Koskela, Ballard, Howell & Tommelein, 2002; Höök &

Stehn, 2005). Lean production can initiate and achieve continuous improvements in a process (Alves, Dinis-Carvalho & Souse, 2012) and has pushed the construction industry towards higher utilization of prefabrication in off-site production (Olsen &

Ralston, 2013).

In literature dating back from 1996 to 2010, the most mentioned drivers for

implementation of off-site production are; more consistent quality and a shorter or more predictable production time and cost (Gibb & Isack, 2003). Despite this, the implementation has been slow (Goodier & Gibb, 2007). One of the main reasons seems to be the high capital cost and the reduced flexibility (Jonsson & Rudberg, 2014).

A solution to this problem might be found in temporary factories. A UK research project called ‘Near Site, Off Site – affordable near site assembly in

(9)

2

Modern Flying Factories’ has been initiated and implemented by Skanska UK, one of the country’s leading construction companies. The research project uses temporary factories, called ‘Modern Flying Factories’ (MFFs) that minimizes required capital investment, which in turn offers increased flexibility. The concept is, at the time for writing, in a phase of performing pilot projects and the company is trying to develop a strategy for implementation in the construction industry. Because of this novelty, the MFF concept is however poorly explored and lacks a defined strategic positioning.

Currently, there are some concepts similar to the MFF concept. Swedish field workshops (SWF) are temporary factories on or close to the construction sites, performed in Sweden during the early 90s and studied by Eriksson (1995). Another, more recent example of temporary factories for the construction industry is the conceptual Flexible Field Factory (FFF). The concept is inspired by the manufacturing industry’s concept of Mobile Manufacturing (MM), which strive for customized production with automated mobile production capacity (Stillström & Jackson, 2007).

To improve the possibility for the MFF concept to be a way for the construction industry to become more efficient, research on the concept is needed. Today, the definition, applicability and potential are not yet studied and such understanding is fundamental for finding its strategic positioning and for further development. Nor is it studied how the MFF concept can benefit from experiences of similar concepts (SWF, FFF, MM) or the application of manufacturing principles (e.g. Lean production, DFMA). In addition to facilitating in the evaluation of the concept’s strategic positioning, such studies might also pose valuable opportunities for increasing the concept’s efficiency.

1.2 Aim and research questions

The purpose of this study is to contribute to increased understanding of the MFF concept, which can facilitate in assessment of future strategic development. The

increased understanding will also serve as a theoretical basis for further research of the concept’s definition, applicability and potential. Since the concept misses a conceptual definition the aim is to propose a theoretical framework that gives a description of the concept based on propositions as well as suggest lessons for its strategic and

theoretical development. To fulfil the aim, the study will answer the following research questions:

(10)

3

RQ1: How can the MFF concept be described?

RQ2: What lessons can the MFF concept learn for future implementation?

1.3 Scope and limitations

The study and the data collection have been conducted in a period of five months, while the MFF concept was in a developing stage and performing pilot projects.

This study is performed on two pilot projects, which so far are the only ones that have been executed as MFFs within the MFF research project. They have been applied to buildings in the UK construction industry and so, the study proposes a theoretical framework for that context. However, the authors have previous experience about the Swedish construction sector. The study is therefore mainly looking at the concept from the UK and Swedish construction industry’s perspective.

The description (RQ1) of the MFF concept is concerning characteristics and positioning in the construction industry. When looking at possibilities for the concept to learn (RQ2), the study is limited theoretically to look at the concepts SFW, FFF and MM as well as Lean philosophies and DFMA. The first ones were chosen because of their similarity to the MFF concept and the second ones were chosen because of their possibilities for creating improvements in a production process.

The purpose of increased understanding of the MFF concept is not specifically addressed to a certain actor in an MFF project. Therefore, the study proposes a theoretical framework of the concept from a broad perspective.

(11)

4

2 METHODOLOGY AND DATA COLLECTION

2.1 Methodology

In the execution, the research was based on a literature review and a case study on the concept with its two pilot projects through interviews and observations, see Figure 3.1a.

The pilot projects produced off-site in MFFs for the construction industry. As such, they offered a good opportunity to explore and evaluate the utility and potential of the MFF concept in the construction industry. The first pilot project in the case study was executed and completed in 2012 and consisted of a four-classroom school extension with prefabricated straw bale panels. The second pilot project was, at the time for writing, in its start-up phase (since October 2014). It is meant to produce 540 utility cupboards as a subcontract for the MEP (mechanical, electrical and plumbing) fitting for apartments. Because of that, the second pilot project provided the opportunity to study the start-up process and the early stage of on-going production of a MFF, while the first pilot project provided some results and evaluations from its use.

The University of Reading was also involved in the research project and in that way, the authors got introduced in the second pilot project. It meant that the authors could observe the actual production and utilization of the concept. That in turn led to opportunities to obtain information from interviews and documents. The latter

consisted of a project funding application and an institutional journal article. Data from these sources dealt largely with the research project and the MFF concept as a whole.

As such, the case study consists of information about the specific projects as well as details of general character.

(12)

5

Figure 3.1a The research methodology. A literature review and a case study of the two MFF pilot projects will contribute to a theoretical framework of the MFF concept

By using a case study approach, the opportunity was given to examine how and why the MFF concept was applied to the pilot projects (Taylor, Dossick, & Garvin, 2011). As the concept is novel, these questions were considered fundamental for being able to describe and learn from the concept. Because of the novelty, the research has had a qualitative approach. The two pilot projects were chosen to serve as a basis for a

deeper understanding (Yin, 2007). As such, surveys and experiments were not suitable, as they were considered too limiting to reach that understanding. Instead, the case- study approach has meant that similarities and patterns that were not expected but found during data collection could be used to create a nuanced picture of the concept (Taylor et al., 2011).

As previous research on the MFF concept was limited, the study has had a mainly inductive (Lancaster, 2005) or exploratory research approach (Taylor et al., 2011).

Instead of having a predetermined theoretical foundation before data collection (e.g.

deductive approach), the theory has been developed simultaneous as the data collection.

During the early stage of the study, it was focused on trying to explain a general

concept of temporary factories. This was to be done by using the MFF concept as a base for comparison with the similar concepts of temporary factories (SFW, MM and FFF) to create a common theoretical framework. However, during the literature review, it became clear that because of a lack of theory around the similar concepts of temporary factories, the foundation on which the research would be built on would be too weak.

Also, the findings during observations and interviews supported a study of the MFF concept on its own better than it would have supported a generalizable theory of temporary factories. As such, it was instead decided to focus on studying the MFF concept and use the other concepts for comparisons in common areas and for lessons to further develop the MFF concept.

LITERATURE REVIEW

CASE STUDY Documents Pilot project 1

Interviews

Pilot project 2

Interviews Observations

(13)

6

The literature review and the theoretical elements which where studied where altered and focused to explain the empirical data that has been found. In order to remain focus, initial research questions served as a guide that the study was directed against.

The preliminary research questions were at the beginning;

1. How can the concept temporary factories be defined?

2. For which projects in the construction industry are the concept of temporary factories suitable?

3. What can the concept temporary factories learn from the mobile manufacturing industry?

4. What are the possible difficulties when starting and operating temporary factories and can lean/5S methods be used for improvement?

The preliminary research questions were revised during and after the late stage of data collection to suit the findings and ultimately support an in depth study of the MFF concept instead. This resulted in RQ1 and RQ2, a propositional description and lessons for future implementation. As such, the inductive approach has helped to guide the study to more reliable conclusions. It has also supported the study’s aim of increased understanding of the concept by formulating propositions framing MFF, since the concept is novel and the theoretical field is poorly explored.

2.2 Methods

2.2.1 Literature review

The research questions required a literature review to effectively utilize existing knowledge around similar concepts, improvement methods and how learning affects production. The literature review included both current and older literature on the subject and was primarily based on peer-reviewed publications. Search was conducted mostly by Google Scholar, PRIMO (Luleå University of Technology’s E-sources), SUMMON (University of Reading’s E-sources) and CIS (Construction Information Service). Examples of keywords that were used separately or combined were; off-site production, mobile manufacturing, flexible manufacturing, temporary construction, DFMA, concurrent engineering and lean production.

Theoretical elements that has been considered relevant and how it has been applied on the different research questions in the analysis is shown by Figure 2.2a.

(14)

7

Figure 2.2a Analysis model of how the theory has been applied to the research questions

2.2.2 The case study

The pilot projects have a central role in the development of a theoretical framework for the MFF concept. They have served as a base in which observations and information from interviews has been compared against visions and literature. Although the concept is not fully developed, it has provided a picture of the concept's current state and helped to evaluate possible strategic directions and development work that the concept can undertake.

Observations

Observations were carried out at the temporary factory in the second pilot project at six times by both authors. Two of the visits had durations of a full working day while four of them were half days. This was done to gather information that helped map the layout and production process in the factory. The authors got introduced to the project when it had already been started, approximately 5 weeks into the production process.

It meant that a lot of the observations also focused on mapping the issues to get a history of the production and understanding of the improvement work.

The observations have been both direct and participating (Yin, 2007). At the first visit, the authors studied the on-going production objectively to get an initial insight to the production. During the later visits, the authors participated in the production by performing quality checks on the cupboards. That meant a greater insight to the design and process of assembly, which facilitated the understanding of the problems and improvement work conducted in the factory. As pointed out by Yin (2007),

participation can result in bias as the objectivity may be reduced. However, since the collected information mostly was of a technical nature, the participation of the authors was not considered to harm or affect the information substantially. All observations

RQ1. Propositional description

Off-site production

Other temporary and mobile production strategies and automation

RQ2. Lessons leanrned

Off-site production

The start-up of a system and learning by repetition

Improvement methods

(15)

8

were complemented through informal interviews with involved people on the sight.

The data collection was performed both during on-going production as well as during breaks.

Interviews

As the concept is novel, the study was dependent on information through practical experience of the concept. Therefore, interviews were held with five respondents who have or have had a position in the pilot projects. This was considered to strengthen their ability to give an overall picture of the concept. The choices of respondents were done together with the supervisor, which had insight into the organisations behind the research and pilot projects. Ultimately, the choices were based on covering different levels of the organisations as persons with different levels of responsibility contribute to different perspectives. That was considered to result in as broad picture of the concept as possible within the study’s time limit. The respondents had the titles of Managing Director for SRW engineering services at Skanska, Sustainability Manager at Skanska, Site manager hired by Skanska, Business improvement specialist at Exelin and Founding Director at White Design Architects. Their involvement in the pilot projects are illustrated in Table 2.2.1 and relations can be found in Figure 2.2b.

(16)

9

Table 2.2 Presentation of the projects and their involvement in the research projects and/or pilot projects.

Respondent Performed Involvement

Managing

Director 19/01/2015

Managing Director at SRW engineering services.

Is involved in the strategic development of the MFF concept.

Skanska

Sustainability Manager

30/11/2014

Involved in the development of the MFF concept and got introduced to it during the start-up of the second pilot project. Is both a participant in the

research project and involved in the management of the second pilot project.

Skanska

Site Manager

24/11/2014

Recruited site manager for the off-site factory in the second pilot project. Does not have any direct involvement in the development of the

MFF concept.

Skanska

Lean Consultant

30/11/2014

Business improvement specialist at the manufacturing advisory firm Exelin. Was involved

in the lean improvement work at the first pilot project and has also had indirect involvement in

the second pilot project by being participant in the research project.

Exelin

Design Director

30/11/2014

Founding Director at White Design Architects.

Has had indirect involvement in both pilot projects by being participant in the research

project.

White Design

Figure 2.2b The interviewed and their relations to the research project and the pilot projects.

The interviews where semi-structured so that the possibility for responses were open- ended. That led to valuable information both in terms of facts, but also in terms of the respondent’s opinion (Yin, 2007). According to Lancaster (2005), group interviews can

MFF RESEARCH PROJECT

PILOT PROJECT 2 PILOT PROJECT 1

(17)

10

get fuller information. However, the risk that the respondents would affect each other's opinions and that only the dominant individuals would get heard was regarded as too high. The interviews were therefore held individually and face-to-face. They lasted between half an hour to two hours, depending on respondents' available time and were conducted at secluded rooms at the university, at the factory site and at Skanska’s head quarter.

The basis when creating the interview questions were observations from the second pilot project, theory from the on-going literature review and any previously completed interviews. The questions where adjusted to the person being interviewed depending on the background, involvement and position in the development or execution of the concept, see Appendix C-G. Depending on findings through the other data collection methods, the interview questions were altered during the process. That was to make sure that the limited time for interviews were used effectively to answer the research questions. Nevertheless, they covered either one or both of the pilot projects as well as the concept as a whole with the main subjects being; purpose, characteristics, execution and future. The questions were designed to cover facts, experiences and personal opinions around the concept.

2.3 Analysis

The analysis was done in accordance to Grounded theory, which is based on the inductive research approach (Lancaster, 2005). The empirical data from the different data collection methods have been analysed separately and merged to create sections of the findings in areas found relevant for the concept. Analysis of the results has been made along the data collection through internal discussions about the findings

between the authors. However, the analysis through processing of the empirical data into the results was done at the end of the gathering. This was done to get a further understanding of the concept when performing the analysis.

The analysis started by colour coding the interview transcripts and internal

documents. As such, parts that were considered interesting for answering the research questions were coloured with the specific question’s colour. Still colour coded, the parts were thereafter rewritten to a describing text with as much of the respondents original formulated words left as possible to minimize faults by interpretation. This was done separately by the two authors and during a review, the parts that were jointly considered important were taken forward to constitute the findings. Making the initial analysis separately was done to screen out irrelevant information and reduce the impact of personal opinions (Lancaster, 2005; Taylor et al., 2011). The findings were then sorted into topics, which the authors considered suitable. These topics form the headings of the result section.

(18)

11

The analysis of the observations from the second pilot project was made in two stages.

The first one was to identify previous or current problems and challenges in the production process and the cause and consequence from them. These findings were then summarized and is presented in the results as a matrix, see Table 5.3, the full matrix for analysis can be seen in Appendix B. This was being done for the purpose of classifying the problems and challenges so that improvements could be identified and suggested as lessons. The analysis was made by both authors in collaboration through a review of the observation protocols. What was considered a problem or a challenge for the production was based on what had been expressed by the employees at the factory or what the authors observed hampered or hindered the production process.

The other stage of analysing the observations intended to identify findings that described the project and production process, which would help answer the research questions. As with the transcripts and internal documents, these notes were also found and selected by the respective author through colour coding and thereafter jointly reviewing. The findings were then merged with the other findings under the appropriate existing or under a new heading.

2.4 The study’s validity and reliability

A presentation of validity and reliability as described by Yin (2007) is presented below.

How they have been accounted for in this study by guidelines from the author will also follow.

Construct validity is treated mainly in the data collection and is about designing real operational dimensions of the concepts studied (i.e. if investigating what is claimed to be investigated). It has been considered by the following:

 Triangulation has been an important part of the study's construct validity. It has been ensured through gathering data from multiple sources with different views/approaches (data triangulation) by different methods (methodology triangulation) and by two researchers (investigator triangulation). It has helped to triangulate the data by comparing what has been stated for the concept in general and what has actually been done in the pilot projects.

 Drafts of the different sections of the report have been sent to the supervisors for review successively during the study to ensure right measures.

 The interview question sheets were sent to the supervisor for review before the interviews to ensure appropriateness and clarity.

 A summary of the results from the interviews was sent to each respondent together with the transcript to ensure correctness.

(19)

12

Internal validity is the establishment of causal relationship and is treated in the analysis of data. Yin (2007) states that this kind of validity does not apply to exploratory

research. However, Newman and Benz (1998) do not agree and instead believes that it is important for every type of research. As such, this study has taken measures to strengthen the internal validity by:

 During the study, alternative explanations have been taken into account as the empiri has been set against each other whenever possible.

 Data has been gathered in a flexible way. Although the analysis of the data was made in the final stage, the data collection has been based on what the already collected data has suggested (i.e. theoretical sampling). For example, data from observations have been part of the basis for the interview questions. In turn, the interviews have been part of the basis for studying some of the internal

documents. It has been an attempt to capture the best theory to explain the MFF concept (Newman & Benz, 1998).

 In addition, the two researchers have been able to contribute with different points of view, which reduces the risk of errors from personal opinions.

 As there exists no previous studies (as the authors know of) on the MFF concept, pattern matching has not been an alternative for ensuring internal validity.

External validity is a delineation of an area within which the study's results can be generalized (i.e. if the results can be applicable in other settings). In this study, it has not been a purpose to generalize the results to a wider context e.g. other concepts of temporary factories. It aims instead to provide a theoretical framework of the concept in its current stage. This framework can then be used by others as hypothesis for further studies, which can replicate the findings to develop the theory even further.

However, Yin (2007) emphasizes that even an exploratory approach contributes to a stronger base for theory building if studying multiple cases. That the MFF concept is novel and has only been applied to the two pilot projects, has therefore implied a limitation of the study.

Reliability deals with whether the study’s design can be repeated with the same results.

In this study, it has been accounted for by:

 The interviews were recorded and transcribed so that information would not be lost on the way and for rendition.

 Findings of interest during observations were photographed and noted

immediately and summaries of the observations were done afterward in a diary as an observation protocol.

 The data is stored on a file hosting service as well as an external hard drive to facilitate retrieval respectively to ensure that the data will not get lost.

(20)

13

 By describing the approach in the study carefully, it is hoped that the study is replicable for other researchers.

(21)

14

3 THEORY

3.1 Off-site production

This chapter review the basis of off-site production and its positioning in the construction industry. How it is defined and how different techniques are categorized will be outlined. The drivers and constraints are reviewed to identify the competitive priorities to the use of off-site production. At last, this chapter also gives an overview of the management and strategy concerns.

3.1.1 Off-site related terms and categorization of techniques

There are different terms and definitions that are associated or synonymous with off- site production, e.g.; off-site fabrication/construction/manufacturing, prefabrication, Modern Methods of Construction (MMC), pre-assembly and standardization (Taylor, 2010).

Prefabrication and off-site fabrication: Jallion and Poon (2008, through Taylor, 2010) defines prefabrication as a manufacturing process, commonly used in a special plant, where a variation of materials is brought together to form a component of the final installation. Further, off-site fabrication is used when both prefabrication and

preassembly are integrated (Gibb, 1999; Jallion & Poon 2008 through Taylor, 2010). The process of off-site fabrication includes design and manufacturing of units or modules, often distant from the construction site. Installation for permanent use is being done at the construction site. Off-site fabrication needs a production strategy that changes the orientation of the process, from construction to manufacturing and installation. (Gibb, 1999)

Modern Methods of Construction (MMC): The term is used by the UK government to describe innovations in the construction industry, which aims to make the house building more efficient. This is done mainly by moving the work from the construction site to a factory where offsite techniques will be used. (Gibb, 1999)

Standardization: Standardization is in literature frequently found together with off-site construction (Gibb, 1999; Goodier & Gibb, 2007; Brege, Stehn & Nord,

(22)

15

2014). Standardization could be defined as “[…] the extensive use of components, methods or processes in which there is regularity, repetition and a background of successful practice” (p. 2) (CIRIA through Gibb, 1999). With careful planning, a high level of standardization and as far as possible in a factory, it is possible to reduce costs and achieve faster construction time (Gibb, 1999).

Categorization and classification in level of prefabrication are made on recurrent similar ways in the prior literature (Gibb & Isack, 2003; Goodier & Gibb, 2007; Brege et al., 2014). A way of classifying off-site construction from a combination of the earlier literature is shown in Table 3.1.

Table 3.1 Categorization of off-site construction in level of off-site (Gibb & Isack, 2003;

Goodier & Gibb, 2007; Brege et al., 2014)

Category Level of off-site

(increasing) Definition

Component

systems 1 Components going into the building.

Example; doors and windows Non-volumetric

units 2 Parts if the building prefabricated in plane elements. Example; slabs and curtain walls.

Volumetric units 3

Parts of the building are prefabricated in modules. Often finalized internal to 'ready to

use' with minimum off-site work. Example;

bathroom pods.

Modular building 4

The whole building is prefabricated in modules. Often finalized internal as far as

possible. Example; housing and prisons.

3.1.2 Drivers

Several articles have studied and in different way, ranked the believed advantages and drivers to implement off-site production as considered by actors in the construction industry (Gibb & Isack, 2003; Venables, Barlow and Gann, 2004; Goodier & Gibb, 2007). Time, cost and quality are clearly the words that are in the top of advantages and drivers (Gibb & Isack, 2003). However, time, cost and quality are terms that can contain multiple meanings, be divided into subcategories and are often related to each other. Example of other common perceived benefits are; increased productivity, minimization of on-site operation, improved health and safety, greater predictability (Gibb & Isack, 2003), reduced reliance on skilled workers, control in production

process and non weather dependent (Venables et al., 2004). Clients are influential in the decision-making and therefore, Gibb and Isack (2003) focused on the client’s

perspective on pre-assembly by interviewing clients. They found that the client’s experience from off-site production where both good and bad. Some had experienced

(23)

16

direct benefits, while others had experienced bad quality and a poorly functioning supply chain.

 Time was the most frequently mentioned benefit by the clients. Within the time benefit it was pointed out that pre-assembly leads to less time spent on site i.e.

speed of construction. Other recurrent time benefits were speed of delivery and less time spent on commissioning. Time is important for clients as the

consequences for not completing on time can mean significant costs. Also, the risk of a change in the market increases with an extended construction time.

 Quality was the second most mentioned category of benefits. A more consistent quality and better control over the quality were frequently mentioned as well as minimizing faults and rework.

 Cost was the third most mentioned category, with benefits such as reduced total and overhead costs through less time spent on site as well as less damages and waste. Cost benefits are important to the clients, especially for repeat-order clients. It was emphasized as important to look at the total cost instead of specific building components. When looking at the total cost, benefits such as decreased preliminary costs, less usage of scaffoldings and get the building weather proof quickly, should be taken in to consideration.

Although time (first), quality (second), and cost (third) are the three biggest reasons for clients to choose prefabrication, quality and cost are, at the same time, the second and third reason not to choose pre-assembly (Gibb & Isack, 2003). This shows that there are different experiences and opinions about what benefits that can truly be obtained by using off-site production. Productivity, people and process where other less mentioned beneficial categories. (Gibb & Isack, 2003).

Although, Neale Price and Sher (1993) argue that prefabrication has clear advantages on the productivity. A factory environment creates possibilities to apply techniques to study time and methods, from which productivity can be systematically improved.

The following is a list by Neale et al. (1993) of influences that affect the productivity through the possibility to produce in a factory:

 A better work environment through better lightening, cleaner and safer working climate.

 Better methods, an example is possibility to cast the bathroom floor up and down in forms, avoiding doing the drainage levels by hand.

 Improved accessibility, for example the possibility of erecting walls in a later stage

 Improved accuracy and less time spent on planning due to that the work becomes more repetitive

 Semi-skilled workers can be trained to manage a limited part skilled-tasks

(24)

17

 Solid factory and labour provides good long-term motivations through secured jobs

 Possibility to reduced movement of workers between tasks and breaks

 Good knowledge of materials and components used regularly

 More efficient sequencing of work by trades operatives

 Working methods can more easy be studied to be improved in detail

 Less damage caused by other operators / actors

 More efficient use of crane, reduced number of lifts

 Easier to introduce specialized techniques and tools

3.1.3 Constrains

Several authors emphasized the decision-making for using off-site techniques as important (e.g. Gibb & Isack, 2003; Goodier & Gibb, 2007). The applicability is depending on the building type, the client market and the financing (Venables et al.

2004). Neale et al. (1993) mentions project characteristics that are important for decision making about using off-site production; planned project duration, income generation of the building in completion, requirements for quality, lack of site working space and difficulty to access. Neale et al., 1993 argue for the number of possible repetitions as clearly a factor to take into account to determine if it is able to respond financially towards traditional construction. The authors also present some project characteristics that make prefabrication principles not suitable:

 In cases where design and production take place at the same time or have significant overlap in time. This means that the design cannot be freeze, which makes it more difficult to get the lead-time needed.

 Volumetric units can be a problem to integrate in the rest of the building because of design requirements

 The transportation to the construction site should not be underestimated. In some cases this may make projects unviable.

The reasons when preassembly (manufacture and assembly usually off-site) did not meet the expectations was mainly because the products where poorly built, that contractors had a lack of experience or that the design was not correct (Gibbs & Isack, 2003).

A typical manufacturer of off-site has limitations that come with the business strategy.

The typical business strategy is to minimize time variations, have a continuous flow in the production and aims to be able to plan the production in a long term. This results in less flexibility and causes difficulties for housing developers or other clients to adapt the techniques, because of a need for more flexible construction methods. That means flexible in being able to react quickly to the demands in both speeding up and slowing down. (Venables et al., 2004)

(25)

18

Blismas, Pendlebury, Gibb and Pasquire (2005) states that implementation of off-site is often hindered by procurement constrains. Procurement constrains presented by Blismas et al. (2005) are when;

 project team members lack experience,

 key decisions are already made that preclude off-site production,

 limited to a particular supply chain,

 obligation to accept lowest cost rather then best value and

 early construction/manufacturing expertise is not available.

The author's refers to the clients having an important role for creating conditions that favour instead of hinder off-site production. Pasquire and Connolly (2002) argue that pre-assembly should be a requirement already in the specification and works best if led by the client.

3.1.4 Labour skills and innovation

Venables et al. (2004) argue that the UK construction industry have in history had a low level of education in comparison to other countries and industries. The UK

construction industry has a market with a turnover of labour that makes it hard to recruit, train and keep labour with the right experience and skills. The author also argue that an increased use of off-site production can be part of solving the problem of lacking skills in the industry, especially in housing construction. Also, Goodier and Gibb (2007) points out that skills shortage has been recognized as a driver for

implementing off-site production. The authors also reported from a survey that 1/3 of the house builders indicated that it was needed the same or less level of skills by the labour for off-site techniques. One pole indicated that increased level of off-site required lower skilled labour and the other pole indicated that it then required a higher level of skilled labour. Clarke (2002) on the other hand, argues that a skilled and trained workforce is needed also in the use of off-site production. It is especially emphasizes that innovation (which is important for example for module making in the construction industry (Neale et al., 1993)) does not happen without appropriate skills and training. As such, Clarke (2002) points out the connection between skills shortage, self-employment and lack of innovations as an explanation to the low productivity of the construction industry in the UK. It is stated that formal training and skills should be recognized to improve productivity.

Aside from the level of skill are also different types of skills. Venables et al. (2004) found that off-site producers do not demand traditionally skilled labour. Instead they are looking for medium level of craft skills and generally interested in semi skilled or multi skilled rather than a specifically skilled workforce.

(26)

19 3.1.5 Changing of roles

An increased usage of off-site production will have impact of the roles in the

construction industry in comparison to the traditional roles. Vokes and Brennan (2013) argued that the use of off-site production had the most significant impact on higher level jobs such as; Site supervisors, Designers, Architects, Structural Engineers, Planners and BIM Modellers. Vokes and Brennan (2013) argue that these new roles give a need for collaboration and integration between disciplines. It is also argued for a need for knowledge in; project management, (BIM and automated design tools),

planning and design. The authors predict an increased integration, through skills and understanding, between design, construction, manufacturing and engineering

disciplines, illustrated in Figure 3.1a.

Figure 3.1a Ideal mix of ‘front-loaded’ skills offsite (Vokes and Brennan, 2013)

Goodier and Gibb (2007) also argue that this early collaborative integration in the process is important for the possibility to choose the right off-site technique. They also found that this collaboration need to include the supply chain, which should get more attention in integration, education and design flexibility.

3.1.6 Decision making and number of repetitions

Since the span of opportunities are wide from non-volumetric applications to fully finished modular buildings, management actions need to be specific to the project. The need of the client and available technology must be considered. (Gibb, 1999). Authors agree that to maximize the benefits of off-site production requires decision timing and an early decision (Gibb & Isack, 2003; Goodier & Gibb, 2007). The timing is different depending on projects and techniques and when Pan, Gibb and Dainty (2008) asked

(27)

20

100 house builders at what stage the off-site techniques where taken into consideration, 71 % answered that they considered it in an early stage (the ’basic house type design stage’). When the consideration of off-site was made early, it was of volumetric

systems, modular building and some more complex panel systems. Off-site techniques such as components, subassembly and some open panel systems could often be

considered at a later stage in the process. (Pan et al., 2008)

The study of Goodier and Gibb (2007) indicates that there are doubts, at the same time as there are believes among contractors, designers and clients, to whether off-site can be done economical. Several authors agree that there is a problem for decision-making in the difficulty to prove economical benefits in comparison to traditional way of building (e.g. Gibb & Isack, 2001, 2003; Goodier & Gibb, 2007; Olsen & Ralston, 2013).

Olsen and Ralston (2013) has found that the decision makers rarely have access to historical data on costs and times as a basis when making decisions to go with off-site production.

Evidence of economical benefits is needed in order to facilitate decisions on application of off-site production. Articles argue that methods should be developed to make decision makers able to do assessment against traditional built. (Neale et al., 1993;

Goodier & Gibb, 2007). Gibb and Isack (2003) describe cases of re-engineering by pre- assembly where successes were proven. Further the authors argue that the success is often in combination with introduced pre-assembly and standardized procedures, which made it hard to say what really allowed the success. Blismas et al., (2005) states that common methods of evaluation are often to calculate and compare costs for material, labour and transportation. This means that benefits from related costs can be neglected such as site facilities, use of crane and reworks. Also, possible benefits concerning health and safety, management and process in total are hard to get into numbers in a economical evaluation. (Blismas et al., 2005).

The suitable scale depends on several factors and it can be complicated to find the right scale to make benefits. However, Neale et al. (1993) exemplify suitable scales for off-site techniques like bathrooms or cladding panels. When the construction is fairly simple, 20 repetitions could be a minimum where as 50 would provide better economic conditions.

(28)

21

Figure 3.1b The relation between unit cost and repetition (standardization) for precast concrete cladding (Gibb, 1999)

The graph in Figure 3.1b displays the relation between unit cost and repetition for precast concrete cladding. It shows a minimum around 10 units to be cost effective and more repetitions will increase the cost effectiveness. However, it also shows that for bigger numbers than 30, the unit cost will differ and more repetitions will get less essential. According to Gibb (1999), this graph looks similar for other off-site components and manufacturing processes. As such, this graph symbolises the importance of knowing when to apply the standardized or off-site constructed components in the projects.

3.1.7 Lead-time and design

Throughout all off-site construction, a longer lead-time represents a significant barrier in the meeting with the traditional construction (Venables et al., 2004; Goodier & Gibb, 2007). Venables et al. (2004) find a big challenge in the fact that the design processes are different in time compared to traditional design processes, which states that it is

important to take into consideration. Goodier and Gibb (2007) argue that to overcome the disadvantages of a longer lead-time, the decision to use off-site needs to be made early to be able to integrate early in the design process and possibly reduce costs. Höök and Stehn (2005) also points out the initial design decisions as especially important, since the longer lead-times that comes with off-site can often mean changes become more difficult in a later stage. Several articles points out that late changes cause problems and prevent efficiency in off-site construction. (e.g. Venables et al., 2004;

Höök & Stehn, 2005). Höök and Stehn (2005) argue that late changes can be caused by a low loyalty to the design of a product, which will reduce the benefits from off-site. It is important that both clients and employees have knowledge about the consequences of late decisions.

(29)

22

Producing in a factory gives other prerequisites than producing traditionally on site.

Pasquire and Connolly (2002) argue that design consultants have a lack of

understanding how to design for manufacture and assembly instead of traditionally.

The authors further state that the designing is not about “carving off a piece of the work” (p.8) and believe that it can be produced in a factory. Considerations are needed to be able to realise the benefits that a factory environment implies. In terms of off-site techniques and approaches to the design, Neale et al. (1993) describe two main

approaches to module design and construction. The first is to have a quite simple construction technique and an approach that basically moves the work from the construction site to a factory. Thus using the same traditional material, techniques and subcontractors as in a traditional built. The second approach is to develop specialized methods to do specific modules.

3.1.8 Design variation and a strategy for standardization

When offering variation in design, it can be handled in different ways. Venables et al.

(2004) argue that the best way to get economy of off-site techniques in housing is aiming for a design with possibilities to offer variation with a small variation in the production line.

There is a balance between standardization and customization that needs to be considered in choosing strategies. An increased level of industrialization

(standardization) must be weighted against the reduced variety for the costumer.

(Jonson & Rudberg, 2012). Several have studied this balance (e.g. Barlow et al., 2003;

Jonson & Rudberg, 2012; Brege et al., 2014). Barlow et al. (2003) presents a model on supply chain strategies based on Japan’s off-site housing industry.

(30)

23

Figure 3.1c Supply chain strategies with examples from the house building industry (Barlow et al., 2003).

Figure 3.1c shows different supply chain strategies where the extremes are pure standardization and pure customization. These in turn include different design strategies and costumer order points. The customer order point or decoupling point is where the costumer requirements enters and is illustrated by the grey shading. The model is based on continuous production. The decoupling point can be seen as a strategic inventory in the supply chain. It is the point that divides the supply chain in before and after costumer orders. (Hoekstra & Romme, 1992 through Barlow et al., 2003).

3.2 Swedish field workshops

Off-site production in temporary factories has been performed before. This chapter gives a short review over the Swedish field workshops, for which concept have similarities with the MFF concept.

Eriksson (1995) wrote a report about development of Swedish field workshops, with weather shelter as the main focus. He describes that tents started to be used during the 60s, mainly for concrete constructions. Timber frames began to be used during the 70s and was until the 80s developed into prefabrication with higher level of completion. It

(31)

24

also went from lifting blocks by hand to using mobile cranes and tents to secure the climate. (Eriksson, 1995)

A field workshop can according to Eriksson (1995) be anything from construction on a worktable to several labours constructing from several steps in the building process.

The author exemplifies various uses for field workshops in wall elements, concrete slabs, reinforcement stations as well as pipes and electrical work. Further Eriksson (1995) is studying the production of wall elements in field workshops, some examples of erected factories performed are; tents, garages (built in a first stage in residential areas), tents with movable part of the roof to be able to use cranes for lifting and a more advanced type of tent with slabs with heating coils and walls in blocks (possible to disassemble). All of them were used with the purpose to secure the climate and to ease handling, the produced elements where transported on a treadmill.

Eriksson (1995) also studied Skanska’s field workshop in Eklanda, Sweden, for a development of 1400 two to five story houses of different types, see Figure 3.2. The construction period lasted about 10 years, starting in 1991. The factory was located in the middle of the field for the development and measured 13x15 m. It was possible to customize and be flexible for the production with a tent construction. They created a wooden floor 0,5-1,2 m above the ground with integrated rails.

Figure 3.2 Photos and layout of the field workshop in Eklanda (Eriksson, 1995)

The idea was to “[…] create an effective and cost conciseness work environment which emphasizes a good working environment.” (translated) (p. 2). The project was to develop systems for light weighted and appropriate field factories while combining good working environment with good economy. Also quality improvements were considered as benefits from being able to weather and moisture proof the wooden construction. (Eriksson, 1995)

Fundamental thoughts and goals with the project in bullet points are (Eriksson, 1995):

 Create a good working environment from an ergonomic point of view.

(32)

25

 The field workshop was planned to be independent from a stationary crane or loader. Loaner was only supposed to be used for transportation for assembly with the finished elements.

 The design of the factory was supposed to be given a high flexibility for a varied use during a long time.

 The factory should be executed on existing ground without expensive groundwork.

 The factory would be built up of hired equipment to enable fast closedown and later on a fast start-up.

3.3 Mobile production strategies and automation

A more up-to-date version of temporary factory production as opposed to SFW, is today's mobile production strategies with automated production processes. In the manufacturing industry, the concept of Mobile manufacturing is an example, while the Flexible Field Factory is a similar concept adapted for the construction industry. This chapter provides an overview of these concepts accompanied by a brief review of some bases for implementation of automation.

3.3.1 Mobile manufacturing

Demands on the manufacturing industry

The literature on manufacturing presents a necessity to re-evaluate current production strategies (e.g. Mehrabi, Ulsoy, & Koren, 2000; Stillström & Jackson, 2007). To be competitive, Jackson and Zaman (2007) says that it is no longer enough to simply produce with low cost and high quality. Companies are also encouraged to create more flexible (Ask & Stillström, 2006; Manufuture, 2004) and unique production capabilities (Jackson & Zaman, 2007), which can quickly adapt to the changing environment (Stillström & Jackson, 2007). Mehrabi et al. (2000) provide a more concise guidance when they state that manufacturing companies must create production systems that can be easily upgraded, in which new technologies and new functions can be easily integrated. The authors believe that this requires a strategy that enables:

 Rapid conversion - for the manufacture and sales of new product models - of manufacturing systems capacity to market demand

 Fast integration of new technologies and new features in existing systems

 Simple adjustment of varying amounts of products for niche marketing

As a summary, ManuFuture (2004) notes that investing in large monolithic mass production plants just seeking to make profit from economies of scale no longer makes any sense.

(33)

26 The concept of Mobile Manufacturing

The concept of Mobile Manufacturing (MM) has mobility, flexibility and speed as its competitive goals (Ask & Stillström, 2006; Stillström & Jackson, 2007). Stillström and Jackson (2007) define mobility as ”the ability to change between geographically different places with little penalty in time, effort, cost, or performance.” (p. 189) The concept of MM is based on the idea that mobile manufacturing modules provide mobile and flexible production capacity for customized manufacturing. By combining the modules, it creates a complete manufacturing system that in size and production capacity can be adapted to suit the need of different customers. The modules have a reconfigurable manufacturing system, which can be quickly and rationally replaced and rearranged to handle new products and/or new volumes. (Stillström & Jackson, 2007)

Ask and Stillström (2006) claim that the responsiveness of production systems will increase with MM, both in the production process and geographically, making all sizes of companies more competitive on the global market. The geographic responsiveness in which production capacity can be close to the need or the skills, is by Stillström and Jackson (2007) emphasized as increasingly important for companies in the future.

Jackson and Zaman (2007) have determined that implementation of MM will enable and support methods for increasing flexibility, mobility and speed, such as:

 Increasing the level of automation in manufacturing processes

 Working with standardized part families by increasing the modularity in products

 Having prefabrication of sub-systems close to final assembly

 Creating close collaboration with sub-contractors and -suppliers

 Developing and improving logistics

When implementing a MM system, it is important with a business model that supports the change as well as efficient knowledge and information handling system. This means that staff that will manage and perform production must be well trained. This is due to that the concept will imply new challenges on a company’s organization,

technology, company strategy, knowledge feedback and information flow. Also, the existing main hard- and software has to be configured for the new requirement and existing methods and systems that support the change has to be identified. (Ask &

Stillström, 2006)

Factory-in-a-Box: Demonstrators of the concept

To demonstrate the concept of MM, five demonstrators have been produced in a project called ‘Factory-in-a-Box’. The project was collaboration between several manufacturing companies in Sweden and four Swedish universities. (Stillström &

Jackson, 2007) The aim of the project was to provide solutions for mobile and flexible

(34)

27

production capacity on demand (Ask & Stillström, 2006; Jackson & Zaman, 2007). In the project, the manufacturing modules consisted of standardized containers that enabled easy and quick transportation to and/or within a manufacturing site, with truck, train or boat etc. (Jackson & Zaman, 2007). They were built for various types of classical manufacturing such as assembly of components or casting of goods. The modules had varying levels of automation, from completely manual to fully automated and the systems were reconfigurable with movable robots and equipment (Jackson &

Zaman, 2007; Stillström & Jackson, 2007). One of the demonstrators where made in order to be transported easily between countries (Jackson & Zaman, 2007). This was facilitated by having a technical solution with a collapsible container, see Figure 3.3a (Stillström & Jackson, 2007).

(35)

28

With this type of production solution, the company was given the opportunity to offer their own production instead of outsourcing it (Stillström & Jackson, 2007).

3.3.2 Flexible field factories in the construction industry

A research project that has many similarities with the MM concept is ‘Flexible Field Factory’ (FFF). Research on the concept has been conducted by Martínez et al. (2013) and involved a conceptual mobile production strategy for the construction industry.

The goal with the study was to create a building-manufacturing paradigm for an open system of products and components through manufacture in factories and on-site. The study also aimed at combining design with highly efficient industrialised production by integrating robotics, ICT systems and new material and technologies amongst others, to the concept.

Within the study, a mobile, semi-automated and reconfigurable factory prototype was simulated for on-site prefabrication of installation modules into an apartment building.

The factory housed in a standard container with integrated legs that allows easy transport and unloading by truck, see Figure 3.3b (Martínez el al., 2013).

Figure 3.3a The technical solution of one of the demonstrators as a combined collapsible container (Stillström & Jackson, 2007)

Modern Flying Factories i byggindustrin: En beskrivning av konceptet och lärdomar för fortsatt utveckling

Modern Flying Factories in the Construction Industry

A Description of the Concept and Lessons for Further Development

Sanna Haukka Magda Lindqvist

2015

ABSTRACT

SAMMANFATTNING

FOREWORD AND ACKNOWLEDGEMENT

EXPLANATION OF TERMS

TABLE OF CONTENTS

1 INTRODUCTION

2 METHODOLOGY AND DATA COLLECTION

MFF RESEARCH PROJECT

3 THEORY