Design Principles and Prescriptions for Planning and Controlling Engineer-to-Order Industrialized Building Systems

Abstract

Construction projects have a high degree of complexity due to both the high degree of uncertainty in process and goals, and the large number of components and stakeholders’ interdependences. In the case of Engineer-to-order (ETO) building systems, there are other sources of complexity, including short lead time, uncertainty related to design, and interdependences between production units. Previous research efforts on the management of ETO industrialized building systems have been limited to managerial improvements from the perspective of companies in charge of manufacturing and assembling components. However, the literature is still scarce on the management of several industrialized building systems that need to be integrated during site installation, considering the perspective of a construction company in charge of the construction stage. The aim of this paper is to propose a set of design principles and prescriptions for production planning and controlling projects that combine different industrialized building systems, considering the key role played by ETO systems in that context. Design Science Research was the methodological approach adopted in this investigation. The development of this set of design principles and prescriptions was based on a literature review and also on an empirical study carried out in a construction project. The outcomes of this investigation are summarized in a framework that establishes interconnections between design prescriptions. The main contribution of this investigation is the development of prescriptive knowledge that can be used to support the design or assessment of planning and control systems that address the requirements of ETO industrialized building systems.

1. Introduction

Construction projects are invariably complex due to the large number of stakeholders involved in the design, delivery, and installation of interdependent components and systems [1] and the high degree of uncertainty in processes and goals [2]. This complexity strongly affects site installation when simultaneous operations are carried out [3].

The adoption of Engineer-to-order (ETO) industrialized building systems has contributed to increase project complexity even further due to several factors: (i) it is often used in projects that have a short lead time, requiring some degree of overlapping between production stages, which contributes to increase the number of interdependencies between tasks [4]; (ii) frequent negotiations with client organizations lead to design specification changes after the start of the design and production phases [5]; (iii) unexpected conflicts may arise among different ETO building systems that need to be installed on-site [6]; (iv) resources such as manufacturing plants and assembly equipment must be shared among different construction projects, increasing the need for efforts in logistics and coordination [7].

The complexity in such environments is sometimes neglected by assuming a high level of predictability during planning and control [8,9]. In fact, management-as-planning is a managerial approach often adopted in construction projects [10], which means setting goals based on long-term estimates of demand and performance, and control is based on checking adherence to those goals [11]. By contrast, the management-as-organizing approach seems to be much more suitable for construction projects that involve ETO industrialized building systems: managers must learn from production to precisely define the following goals [11], and plans should be refined based on information that emerges from production [10].

The lack of communication among stakeholders is an important issue in the management of those systems. Decision-making is often fragmented, and plans are frequently push-driven from upstream to downstream production units along the supply chain [12]; e.g., long-term plans are used for triggering the production and the delivery of components. These problems are often caused by procurement decisions [13], e.g., contracts often link payments to the production or delivery of components rather than to site installation.

Rauch et al. (2018) [14] stated that the uncertainty can be managed to some extent by introducing short re-planning cycles based on regular feedback from the on-site installation status. This idea is aligned with the extended conceptualization of pull production proposed by Hopp and Spearman (2011) [15]: “what distinguishes a pull from a push system is the fact that the former work is released according to system status, rather than on estimates of demand.” This definition is relevant to the context of ETO industrialized building projects, as pull production can be regarded as a mechanism for coping with uncertainty and controlling work-in-progress (WIP) [16,17,18].

Previous research efforts have identified principles, requirements, or recommendations related to planning and controlling ETO industrialized building systems, some of which are related to supply chain coordination [4,19] or to construction site logistics [17,20]. However, those studies have focused on improvements in individual ETO systems, taking the perspective of companies in charge of manufacturing and assembling components. The literature on the management of several ETO industrialized building systems that need to be integrated during site installation, considering the perspective of construction companies that manage the whole construction stage, is still scarce [21]. In this context, planning and control systems must fulfill additional functions compared to traditional building projects due to the need to coordinate different types of flows in each project stage, i.e., design information, fabrication, logistics operations, and site installation [21,22]. Against this background, the research question addressed in this study is: how to design production planning and control systems for projects that involve multiple ETO industrialized building systems?

The aim of this paper is to propose a set of design principles and prescriptions for production planning and controlling in projects that combine different industrialized building systems, considering the key role played by ETO systems. This research work explores the synergies between Lean Production principles and BIM functionalities, as suggested by Sacks et al. [23], to support the integrated management of industrialized building systems. Moreover, differently from previous studies, the focus of the implementation of Lean Production is on managing production in single construction sites, but considers the need to integrate different production units by managing flows between them [24].

A design prescription can be understood as a suggestion for action in a given circumstance to achieve an effect [25], also named functional rules [26]. Design principles can be regarded as categories of design prescriptions, i.e., recommendations that support decision-making in design [27]. Both design principles and prescriptions must support decision-making for classes of problems [28] and should provide a coherent conceptual framework for scholars and practitioners to understand each other [29]. The outcomes of this investigation are represented in a framework that establishes interconnections between design prescriptions. The development of this framework was initially based on a literature review, and then the inter-relations among design prescriptions were defined, considering an empirical study carried out in collaboration with a construction company.

There is a strong connection between the outcomes of this investigation and the performance of construction projects in terms of sustainability, considering the environmental, economic, and social perspectives [30,31]. The dissemination of industrialized building systems can be regarded as an important strategy for achieving sustainable construction by bringing several benefits, such as the increase in productivity [32], improvement in safety and working conditions [31], short delivery time [33], higher reliability in terms of quality [34], and energy and waste reduction [35,36]. Dallasega and Rauch [30] suggested that the positive impact of industrialized construction on sustainability is directly related to the implementation of some core production management concepts and principles, such as limiting the amount of WIP throughout the supply chain and synchronizing processes carried out by different stakeholders.

This paper is structured in six sections, including the introduction. The Section 2 presents core concepts related to ETO production systems. The Section 3 presents the research method, which is positioned as design science research. Then, the results of the empirical study and the framework that establishes interconnections between design prescriptions are presented in the Section 4. Section 5 presents the discussions, and finally, the main conclusions and opportunities for future research are depicted in Section 6.

2. Literature Review

The literature review is divided into three parts. The first one presents some core concepts related to ETO production systems, including flexibility, modularity, slack, and flow, followed by a compilation of design principles and design prescriptions for ETO industrialized building systems. This discussion provides evidence that some of the proposed principles and prescriptions used in this investigation have already been discussed in the literature, but this knowledge is still very fragmented. In the third part, three production planning and control methods often associated with the implementation of the Lean Production philosophy in construction are introduced, i.e., the Last Planner System^® (LPS), Location-based planning methods (LBP), and Four-dimensional Building Information Modeling (4D-BIM). Those production planning and control methods have been effective in supporting the management of complexity in construction projects by addressing the shortcomings of methods based on the management-as-planning approach.

2.1. ETO and Related Concepts

Viana et al. [20] and Powell et al. [37] addressed the importance of product flexibility for companies that deliver ETO industrialized building systems by adapting products to fulfill specific customer requirements [38]. Additionally, ETO production systems need to be flexible in terms of volume, mix, or process due to the high degree of uncertainty in demand [8].

Modularity is a concept that plays a key role in the development of flexible production systems [39]. According to Gershenson et al. [40], it is concerned with the idea of dividing a product into manageable parts (modules) that have standard ways of interaction. Modularity is strongly related to the core operations management principle of reducing the batch size, which can be related to product modularity (e.g., a single sub-assembly) or process modularity (e.g., a small number of interdependent processes that are necessary for installing similar sets of components) [41].

Both uncertainty and flexibility can be managed by introducing different types of slack [42], which can be defined as a cushion of actual or potential resources that allows an organization to adapt successfully to internal and external pressures [43]. Slack strategies can be classified into two core categories: (i) redundancy: resources are provided in addition to the minimum necessary; (ii) flexibility: resources that can be used in different ways, e.g., multi-skilled workers and multi-purpose equipment [44]. Therefore, slack is a broad concept that can be used in a wide range of situations, both from a technical or organizational perspective. For instance, slack can be used to fulfill demands or perform actions at a higher strategic level, such as engaging organizational participants, conflict resolution, and innovation [44]. Some types of slack have an opportunistic character, due to the capacity of human actors to change by self-organizing [44]. Buffers is a more specific concept, and can be regarded as a kind of technical slack, often included in production plans and cost estimates [45]. Moreover, buffer management techniques usually deal with variability statistically described in advance, with a limited application when uncertainty is difficult to anticipate and quantify [46].

Powell et al. [37] highlighted the importance of systematic cross-functional and inter-organizational integration in the management of ETO manufacturers. Wiendahl et al. [47] emphasized the importance of defining consistent objectives and communicating responsibilities among supply chain members to have individual and shared goals for guiding production planning and control. From the perspective of construction projects, project delivery systems must have governance mechanisms that support the alignment of stakeholders’ goals (e.g., client, contractor, and suppliers) [13].

The importance of managing flows in ETO industrialized building systems is also pointed out by different authors [20,48,49]. However, as suggested by Shingo [50], it is necessary to standardize and manage both product (e.g., prefabricated components) and operations flow (resources such as shared equipment or assembly team). Therefore, logistics management plays a key role in the management of ETO industrialized building systems. Different activities are involved: (i) coordination of manufacturing plants, delivery of components on-site, and site assembly; (ii) design of loading and unloading operations; (iii) definition of site layout, including the location of temporary facilities, equipment, storage areas, and pathways; (iv) decision-making in situations related to space or time conflicts [17,51].

As previously mentioned, this type of production system introduces other sources of complexity due to the need to integrate different production units and supply chain members that assume new roles [52]. As a result, the overall construction process tends to become more complex due to the combined effect of interdependence between off-site and on-site processes and uncertainty caused by internal and external factors [53,54,55]. In this context, information-sharing practices should give stakeholders access to more accurate information on inventories, WIP, flow rates, and orders throughout the supply chain, making overall control possible and increasing transparency [21].

2.2. Design Principles and Prescriptions for Planning and Controlling ETO Production Systems

Seven design principles for planning and controlling projects that integrate different ETO industrialized building systems were explored. The adoption of ETO prefabricated building systems requires plans to be refined according to status information, which aligns with the management-as-organizing approach [11]. Thus, the first design principle for planning and control systems in this context is to manage uncertainty, which can be associated with two prescriptions. The first is to establish different hierarchical levels of planning and control so that more details are added as more information on the project becomes available. The literature on construction planning and control often suggests three primary planning levels: long-term, medium-term, and short-term planning [56,57]. The second prescription is to design slack by making redundant resources, such as time, capacity, or contingency plans. This is mainly intended to reduce tight couplings to absorb the effects of uncertainty to soften the core characteristics of complex systems [42].

The second principle is to deal with customer-oriented and non-repetitive production and variable routings, initially proposed by Viana et al. [20], which is concerned with considering internal or external customer needs. The application of this principle is related to different types of flexibility, as suggested by Abdelilah et al. [58]. The first associated prescription is to have operational flexibility, which means using adaptable resources, which can also be regarded as a type of slack. It is related to using the same resource to perform different functions, such as a multifunctional and cross-trained workforce. The second prescription is to adopt process modularity. The division of the project into modules can contribute to create repetitiveness despite the fact that some projects have limited product repetition, as well as to increase output flexibility [38].

The third principle is to promote stakeholder and systems integration, as suggested by Powell et al. [37]. The purpose is to avoid fragmented communication and improve the exchange of information among stakeholders, as highlighted by Cho and Fischer [12]. The first prescription is to encourage collaborative work among stakeholders, which can improve interface management, decision-making, and management of trade-offs [21]. It represents an attempt to engage people in managerial decisions [57]. The second prescription is to align stakeholders’ interests and goals (e.g., client, general contractor, and suppliers) through governance mechanisms in order to create a suitable environment for improving collaboration and coordination. The application of some types of project delivery systems that encourage collaboration between stakeholders, such as Integrated Project Delivery (IPD) or Target Value Design (TVD), can be regarded as mechanisms for implementing those two prescriptions [59].

The fourth principle is to plan for flow, i.e., changing how production is traditionally controlled, mainly focused on value-adding work. The first prescription is to define small batches as managerial units to control the flow of components and synchronize production in manufacturing plants and construction sites. Those batches should be sized and designed to fulfill different needs in project stages, such as design, production, transportation, and assembly. As Koskela [38] pointed out, reducing batch size creates opportunities to identify problems and learn and adapt from previous cycles. The second prescription is to manage both product and operations flow. According to Ballard [60], the necessary resources must be available at the right time to have a stable and predictable production system, increasing the production flow reliability. It embraces logistics management, which is related to the management of information and physical flows, i.e., storage, handling, transportation, and distribution of resources [61]. The third prescription is to control work-in-progress. Locations must be used as units to control WIP, reducing the difference between the progress of two consecutive and dependent processes [62].

The fifth principle is to adopt pull production. The production and delivery of prefabricated components in manufacturing plants must be pulled by site installation. Depending on the type of components and systems to be installed on-site, distinct configurations of production systems with different lead times should be considered [33]. The first prescription is to adopt capacity reservation as a mechanism to fulfill the construction site demand, despite the uncertainty that exists in the site installation. Long-term plans can quantify the demand for components produced off-site and reserve the manufacturing plant capacity. Details about each delivery batch should only be provided when more information about site assembly is available. The second prescription is to establish confirmation points at lower hierarchical planning levels, which can be used to inform manufacturing plants about changes in site installation sequence or timing [18]. It is a mechanism to improve the use of the capacity of the production units, avoiding the production of components in advance and preventing overproduction. In this case, long-term plans are no longer the only source of information for scheduling at manufacturing plants. The third associated prescription is to display the system status, which enables pull production to occur. As Gosling et al. [21] highlighted, there must be a frequent and transparent exchange of information between different production units, so the components can be delivered according to site installation demand.

The last design principle is to create continuous improvement, which requires leadership from production managers, and social skills related to communication and management of collaborative processes [37]. Training people from different hierarchical levels also plays a key role so that they can contribute to planning decisions [63]. The first prescription is to learn through feedback cycles, typically part of planning and control systems [57,64]. More rapid feedback cycles allow faster identification of problems and more rapid learning and improvement [65]. The second prescription is to develop the necessary competencies for implementing Lean concepts that are not widely cultivated in the construction industry, such as pull production and slack.

Figure 1 schematically presents the set of design principles and prescriptions mentioned above.

2.3. Production Planning and Control Methods in Construction

2.3.1. Last Planner System^®

LPS is a planning and control method, which is divided into hierarchical levels. It transforms what should be done from the long-term plan into what can be done through constraints analysis at the look-ahead level to form an inventory of made-ready work that can be included in the short-term plan [66]. It is a collaborative and decentralized planning approach that involves trade leaders and lower-level management in decision-making, creating short learning cycles [57]. These short cycles (e.g., one week) support learning by involving different stakeholders in data analysis and planning decisions, based on a set of simple metrics (e.g., percent of plan completed—PPC and causes of non-completion of work packages) [57,64].

LPS can also be regarded as a planning model suitable for the high level of variability that exists in construction projects. There are different mechanisms for systematically introducing slack: (a) resources that are made available in advance by the removal of constraints [67], (b) a mechanism for matching load and capacity at the short-term planning level, which is a kind of capacity buffer [60], and (c) the backlog of made-ready activities, i.e., contingency plans [68].

Based on the definition of pull production proposed by Hopp and Spearman [15], LPS can be considered a combination of pull and push planning. The role of long-term plans is to push completions and deliveries based on forecasts. At the look-ahead and short-term levels, production is pulled according to system status, i.e., availability of resources, matching between available capacity and demand, or emerging interferences among crews.

In summary, the use of LPS could help the general contractor to address the following design principles: “manage uncertainty,” “promote stakeholders and systems integration,” “adopt pull production,” and “create continuous improvement.”

2.3.2. Location-Based Planning Methods

LBP has been used for explicitly planning for flow by showing interactions of construction processes within locations [69]. This method makes it possible to further streamline the workflow by structuring work according to space availability, recognizing that there is always contention for space, such as where work is to take place, materials are stored, and access is needed [66]. LBP can be effectively used to analyze space as a constraint [70]. It requires establishing a project location breakdown structure (LBS), specifying locations where work will take place [69].

LBP allows the direct application of process modularity, defined by Voordijk et al. [40] as the degree to which the production process can be broken down into sequential sub-processes with similar content. Process modularity is also strongly related to the principle of defining small batches as managerial units, which can contribute to WIP control, lead time reduction, and early detection of deviations. This approach facilitates the use of visual tools, including the line of balance (LOB) technique [71].

This approach could support the following design principles: “deal with customer-oriented and non-repetitive production, and variable routings” and “plan for flow.”

2.3.3. Four-Dimensional Building Information Modeling

Building information modeling (BIM) allows the use of a virtual model as a central and consistent source of information for stakeholders [72]. In production management, 4D BIM models are particularly relevant, as they can be used to manage product and operations flow, considering time, resource, and logistical elements, and contribute to the improvement of construction safety [73,74]. These models support the development of accurate plans and communicate effectively key aspects of plans, including construction methods, task sequences, the definition of work locations and batches, and flows of components [75,76].

The visualization of the construction processes through 4D model outputs (e.g., videos and screenshots) can be used to identify spatial constraints by stakeholders. Such models can be continuously updated as construction progresses and the physical characteristics of the construction site change, contributing to displaying the status of the system [18]. Those models are strongly related to the “manage uncertainty,” “plan for flow,” and “adopt pull production” principles.

3. Research Method

3.1. Methodological Approach

Design Science Research (DSR) was the methodological approach adopted in this investigation. This approach is field-problem-driven and solution-oriented, in which alternative courses of action in dealing with sociotechnical systems are devised, resulting in a theoretical contribution [28]. The main outcome of this investigation is a set of design principles and prescriptions for production planning and controlling ETO industrialized building systems. The set of principles and prescriptions is summarized in a theoretical framework that establishes interconnections between them.

As mentioned above, the literature review was the point of departure for the development of the initial version of the artefact. Then, an empirical study was carried out, in which inter-relationships among prescriptions were explained. That empirical study had a holistic character, in which multiple sources of evidence were used, similar to a case study [77], allowing an in-depth data analysis for testing and refining the proposed set of principles and prescriptions.

This research was developed in close collaboration with a Brazilian construction company (named Company A) that has operated for more than 50 years. This company had an ongoing operational excellence program based on the Lean Production philosophy, aiming to improve production management processes and routines in construction projects. The empirical study was carried out in a school expansion project of approximately 55,000 m², located in Porto Alegre, Brazil. Company A was the general contractor, in charge of the whole construction stage, including the coordination of suppliers and subcontractors. In this project, on-site construction activities run in parallel with school operations. For that reason, the site was divided into zone 1 (four buildings) and zone 2 (two buildings), which were tackled at different times, allowing the continuation of educational activities.

3.2. Research Design

The empirical study lasted for ten months, and the research process was divided into three phases: understanding, development, and assessment, as shown in Figure 2. The first phase took three months to understand the project’s context and identify improvement opportunities in the existing production planning and control system.

The sources of evidence used in phase 1 are presented in

Table 1. Sources of evidence adopted in phase 1.

Source of Evidence	From	Aim
Direct observation and photographic records	Thirteen visits to the construction site (around 2 h)	To collect data on general characteristics of the construction process (site layout, flow of components, use of visual devices), details of the existing planning and control system (defined hierarchical levels, metrics for controlling teams’ performance), and to observe site installation processes.
Participant observation	Nine short-term planning meetings (1 h each), involving representatives of Company A and subcontractors	To understand routines and identify improvement opportunities
Document analysis	3D-BIM models, production long-term and short-term plans, and production control spreadsheets	To gain an overview of the design and to understand planning and control tools.
Semi-structured interview	Four interviews: design manager, site manager, site coordinators from two subcontractors-structure and building services (e.g., electric, plumbing, sewage, HVAC, fire detection (around 1 h 30 min each))	To understand procedures for production planning and how information was exchanged among Company A and subcontractors.
Participant Observation	Two training courses (4 h each)	To improve knowledge of basic Lean Production concepts and principles and production planning and control.

. In addition to some data collection activities, an 8 h training course about lean concepts and production planning and control methods was organized for the managerial staff of Company A so that they could better contribute to the understanding of the existing problems and identification of improvement opportunities.

The development and assessment phases were undertaken in seven months. In the development phase, several improvements were introduced in the production planning and control system, similar to what has been described by Sein et al. [78] as action design research. Therefore, there were learning cycles involving interactions between the research team and the representatives of Company A.

The development phase was divided into two sub-phases. The first addresses production planning and control improvements in Zone 1 by refining the LPS, especially regarding the look-ahead planning level, and by introducing LBP and 4D BIM. The second encompassed the development of a long-term plan for Zone 2 using the line of balance technique based on the information collected in the previous zone. The learning cycles involved planning actions, implementing those changes, evaluating the results, and reflecting on the research contributions, as suggested by Eden and Huxham [79].

Synchro PRO (version 6.4.3.2) and Vico Office (version R6.5) were used as software tools for devising 4D models and LOB, respectively. Synchro PRO was chosen mainly because of its functionalities for modeling logistics operations. Vico Office was chosen because it adopts the location-based planning approach as a basis for developing lines of balance. The sources of evidence used in the development phase are presented in

Table 2. Sources of evidence adopted in the development phase.

Sources of Evidence	From	Aim
Direct observation and photographic records	Thirty-one visits to the construction site (around 2 h)	To collect information about the progress of construction activities and compare them with plans. One of the researchers also had the opportunity to engage in informal conversations with workers and managers to understand production issues and causes of deviations.
Participant observation	Twelve short-term planning meetings (1 h each) involving representatives of Company A and subcontractors	Discuss the weekly performance by analyzing quantitative and qualitative data (e.g., PPC, batch adherence chart, causes for the non-completion of work packages), and the weekly plan for the following week.
Participant observation	Fourteen medium-term planning meetings (2 h each) involving representatives of Company A	To support the development of a routine for implementing this hierarchical planning level.

.

The third phase of this research project consisted of assessing this investigation’s practical and theoretical contributions. The results of the implementation process were assessed during four workshops (around 2 h each) organized by the Innovation Department of Company A. In those workshops, the participants performed a critical analysis of the contributions of the empirical study to the planning and control system of the company. That analysis contributed to the development of the framework that explicitly shows the connections among the design prescriptions for planning and controlling different ETO industrialized building systems.

4. Results

4.1. Existing Production Planning and Control System

The existing production planning and control system had only two levels: long- and short-term planning. The long-term plan had over 1000 activities and was produced by using a critical-path-method (CPM)-based software. It was mainly used as a contractual document for reporting project progress to the client organization. The focus of production control was on deliverables and time deviation, as in any CPM-based planning system.

In the long-term plan, the prefabricated structure assembly process was not sufficiently detailed, e.g., one activity of the plan represented the assembly of all building columns. However, this type of system demands several planning and control details due to the interdependence between the production units (manufacturing plant and construction site) and other building systems to be installed. Additionally, the construction site was a confined area, resulting in a single route of materials transportation and limited space for storage, further increasing the need for coordination and logistics management. The agreement between the general contractor and the company in charge of the delivery and assembly of the prefabricated structure established part of the payment based on the volume of components produced and ready for on-site delivery. The manufacturing plant coordinator was in charge of producing load plans and planning the delivery of components to the construction site by considering the capacity of trucks and the demand from site assembly, which was established by another manager.

Short-term planning was carried out in weekly meetings, similarly to LPS, in which the PPC was monitored, and the causes for the non-completion of tasks were identified—the average PPC was 57%. This meeting was overloaded by discussions related to constraint analysis due to the absence of the look-ahead planning level. Therefore, the long-term plan was used to push weekly tasks without considering production status. Earthmoving, pile foundations, drainage systems, and prefabricated structure assembly were the activities being executed during phase 1.

At the end of this research phase, some improvement opportunities in the planning and control system were identified, as shown in

Table 3. Improvements opportunities.

Design Principle	Improvement Opportunities
Manage uncertainty	Develop collaborative and decentralized planning by adopting three hierarchical levels. Introduce the look-ahead planning level. Involve crew leaders and lower-level management in short- and medium-term planning meetings.
Deal with customer-oriented and non-repetitive production and variable routings	Make an effort to capture customer requirements, both internal (e.g., different work teams) and external (e.g., the client organization).
Promote stakeholders and system integration	Improve communication between the construction company and suppliers. Use visual management tools to disseminate relevant information to stakeholders, such as performance metrics and production system status. It is particularly important in planning meetings to increase the degree of understanding by planning participants.
Plan for flow	Use location-based methods and synchronize processes by defining takt times. Create interdependence between wagons (processes to be synchronized). Introduce explicit time buffers in the LOB. Use of 4D BIM models to analyze construction site logistics, including flows of components, access and storage areas, location of temporary facilities, and definition of pedestrian and vehicle traffic routes.
Adopt pull production	Define confirmation points to integrate prefabrication and site installation demands to minimize the effect of variability and to control works-in-progress.
Create continuous improvement	Assess the performance of existing processes and introduce learning cycles.

. Those suggestions for improvement were linked to the design principles previously presented.

4.2. Implementation of Improvements in Planning and Control

Due to the non-repetitive nature of the project, it was necessary to divide the project into batches with a similar amount and type of work for each location so that a takt time could be defined to synchronize some processes. The LOB (Figure 3) was used to represent the long-term plan, and the 4D-BIM model (Figure 4) was used to visualize both product and workflows.

Groups of work packages from the line of balance were planned to be in a continuous flow, similar to wagons in a train. Each set of packages had similar cycle times, slightly shorter than a takt time [80]. The takt time was not the same for all processes, with it being necessary to add buffers between sets of synchronized processes so that these would not interfere with each other. For each group of processes, one was chosen to pull the others so that a similar pace was planned for them. For example, the drywall partition finishing process was selected to establish the rhythm of a set of processes so that three types of ceilings being executed by different subcontractors had to follow the same pace, as shown in Figure 3.

The client organization (end-user) demands in terms of areas to be delivered early in the project had to be considered in the definition of project batches and execution sequences. In Building C, for example, there were classrooms, and the client requested expediting delivery because of the need to install furniture items and electronic equipment, which was not in Company A’s contract scope.

The concrete structure assembly batches were jointly defined with representatives of company A and the company in charge of delivering the prefabricated structure. The average productivity index provided by the supplier of concrete was used for producing the long-term plan: ten columns per day or a set of eighteen components (beams, slabs, stairs, and facade panels) assembled per day. Figure 5 presents 4D images for assembly batches, in which different colors indicate the crane’s position. Logistic decisions had to be taken in parallel with the site installation plans in this project. The 4D-BIM model was used to support logistics decisions on the flow of components, access and storage areas, location of temporary facilities, and definition of pedestrian and vehicle traffic routes, as shown in Figure 6.

The look-ahead planning and control level was implemented, emphasizing the identification and removal of constraints. Weekly planning meetings involved contract managers, production, engineering, and material supplier representatives. The categories of constraints were design, material, labor, equipment and tools, method, and health and safety. Logistics decisions were also discussed at look-ahead meetings in parallel with revising the assembly sequence.

The short-term planning level was also improved. Visual boards were employed to support this level of planning. The work zone control board (Figure 7) was developed to support short-term planning meetings. It explicitly indicated where each team (identified by each number in Figure 7) had to perform their activities and the available areas for material and equipment storage on each floor. Thus, those visual tools helped to improve communication and information exchange between stakeholders. In addition, as suggested by Ballard and Howell [56], the role of short-term planning meetings in the management of commitments was emphasized.

In the second research development sub-phase, a long-term plan for the project Zone 2 was jointly developed by managers from Company A and the research team. Based on the sequence of activities and existing interdependencies, three location-breakdown structures (LBSs) were defined, as shown in Figure 8. The LBS considered integrating the installation process of different building systems and off-site manufacturing and installation. The Vico Office software used in this phase enabled the definition of different location systems for the same project, allowing elements to be decomposed differently depending on task needs. Figure 9 shows the LOB produced for representing the long-term plan, in which all activities are connected to the location breakdown structure.

The average PPC in the first development sub-phase was 69%, representing an incremental improvement compared to the initial situation, primarily due to the implementation of the look-ahead planning level and an increase in the engagement of the stakeholders in both short- and medium-term planning and control levels.

Moreover, visual tools supported decision-making during those meetings, contributing to sharing product and process information. The main cause of the non-completion of work packages was the lack of labor from the subcontractor, representing 56% of the causes. Two subcontractors in charge of masonry and mortar rendering had financial problems. They reduced the number of workers on-site, which caused delays in interdependent activities such as water and electricity pipes.

In the 24 weeks of data analysis of the look-ahead planning, 82% of constraints were removed, and 37% were removed without rescheduling. For instance, the constraint “lighting design details” had to be rescheduled 11 times, delaying the installation of lighting appliances for one month. Another constraint, the technical specification of windows, had its removal rescheduled seven times due to delays in the client’s decision-making. The postponement of the client’s decisions resulted in delays in the fabrication and delivery of windows, which affected the execution of the drywall partition finishings. Those problems indicate that it was necessary to extend the planning and control system to upstream stages, including design and client decision-making.

Figure 10 presents the batch adherence chart specifically for the prefabricated structure assembly. It provides evidence that new assembly batches were started without concluding the previous ones, increasing the amount of WIP and the incidence of non-value-adding activities. Figure 11 shows the comparison between the number of components that were fabricated and installed. The dotted red line represents the remaining number of components ready for installation but kept in the storage space at the plant yard or the construction site.

On average, each component took 27 days from the off-site prefabrication to the final site installation. In addition to the need for an available storage area, the use of plant capacity and production of components in advance resulted in the double handling of materials. This problem reinforces the need for contracts and other governance mechanisms that encourage client organizations, the general contractor, and suppliers to remove constraints on time and synchronize the production of prefabricated components and site installations. With the adoption of large batches in the assembly of the prefabricated structure and the limited effectiveness of the look-ahead planning, many changes in the sequence of downstream activities had to be made, such as concrete floors, drywall, waterproofing, and roof execution.

4.3. Description of the Framework

The set of design principles and prescriptions was revised based on the analysis of the empirical data. Some interactions between the prescriptions were identified, as shown in Figure 12. These interactions represent non-hierarchical relationships and indicate that there is a combined effect of the design prescriptions, indicating that the set of design principles and prescriptions works better if these are simultaneously implemented. Dashed lines connect each design principle with the corresponding prescriptions, while dotted lines represent interdependences between design prescriptions. The framework is explained in the following paragraphs.

“Define small batches as managerial units” is one of the steps that are necessary to “adopt process modularity” and contributes to reduce the cycle time and to “learn through feedback cycles.” In the empirical study, an effort to adopt process modularity was made through the implementation of LBP, by detailing assembly batches (e.g., Figure 5). Those batches were defined in order to enable some degree of repetition, even though the batches were not identical in terms of the components’ mix.

“Have operational flexibility” and “design slack by making available redundant resources” can be applied to “manage both product and operations flow.” By defining sets of interdependent and synchronized wagons in the line of balance, both product and process flows become easy to understand and monitor. Different types of slack are necessary (e.g., time, space, and capacity) in production plans in order to make those flows reliable and reduce interferences during the execution. Examples of slack can be identified in the line of balance (Figure 3): allocation of two teams to execute the same work package but in different locations to reduce the cycle time (blue dotted arrow); work zone available and waiting for workers due to the prioritization of the continuous flow of workers (red dotted arrow).

“Display system status” is necessary to “control work-in-progress” and enables “learn through feedback cycles.” As presented above, visual tools, such as the LOB, 4D-BIM, and the work zone control board, can be used to disseminate product and process information to stakeholders during planning meetings (see Figure 7). These tools also increase process transparency, enabling problems to be quickly identified. The look-ahead and short-term hierarchical planning levels require information about the system status as some tasks need to be pulled.

“Establish different levels of planning and control” and “encourage collaborative work” improve the creation of consistent plans and support “learn through feedback cycles.” In the empirical study, the combination of LPS and LBP can be regarded as a strategy to manage uncertainty by establishing different hierarchical planning and control levels. The long-term plan was produced using LBP methods, while the look-ahead planning level and the short-term planning were improved, contributing, to some extent, to increase the reliability of the production system. Moreover, visual management devices were used to support collaborative work among stakeholders (Figure 7).

“Establish different levels of planning and control” provides information according to the level of detail needed to “adopt capacity reservation” and “establish confirmation points.” The level of planning details increases as more information is made available along the project. The long-term plan (e.g., Figure 3 and Figure 9) can be used initially to reserve the manufacturing plant capacity. It avoids the production of components in advance, preventing overproduction and waiting times. As soon as the installation sequence and timing are confirmed, detailed site installation plans are communicated to the manufacturing plant (e.g., Figure 5). Moreover, confirmation points can be included at the very operational level, e.g., in the daily delivery of components.

“Develop the necessary competencies” supports “align stakeholders’ interests and goals through governance mechanisms.” Training courses were conducted for the company’s managerial staff to make them able to understand and apply core lean concepts related to production planning and control. However, limitations regarding suppliers’ integration in the planning systems were observed, partly due to the inadequacy of governance mechanisms.

“Adopt capacity reservation” and “establish confirmation points” can be used as a strategy to “align stakeholders’ interests and goals through governance mechanisms.” The need to extend the planning and control system to upstream stages (e.g., design and customer integration) and establish confirmation points are strategies that must be used for aligning stakeholders’ interests and goals. ETO suppliers (e.g., prefabricated structure) need more time to design, manufacture, and assemble their products than a supplier that has the products stored and ready for delivery (e.g., bricks supplier), so the lead time for each resource must be considered during the planning process.

5. Discussion

The artefact devised in this investigation is a set of design principles and prescriptions for planning and controlling projects that combine different ETO industrialized building systems. This set was formed by combining different contributions extracted from a literature review and then refined based on the insights from the empirical study. A framework was proposed wherein the interactions between the prescriptions were explained using non-hierarchical relationships. These interactions indicate that the impact of those design principles and prescriptions depends on their combined effect.

The proposed framework is applicable to the specific context of construction companies that need to integrate ETO industrialized building systems in site installation rather than being a set of general guidelines to be applied in any type of production system, such as the principles proposed by Koskela [38], Womack and Jones [81], and Liker [63]. As the perspective of a company that oversees several building systems was considered, the scope of this artefact is broader than the frameworks previously proposed by Viana et al. [20], Gosling et al. [21], and Powell et al. [37]. Those were devised for companies that deliver single ETO systems. Other principles that are relevant for the context of general contractors that need to integrate several systems have been included in the framework, such as “design slack,” “alignment of goals among supply chain members,” and “plan for flow.”

From a practical perspective, the artefact can be used to understand the underlying ideas and the roles of planning and control methods, providing support to assess existing production planning and control systems or to design new ones. However, some prescriptions such as “align stakeholders’ interests and goals (e.g., client, general contractor, and suppliers) through governance mechanisms,” “adopt capacity reservation,” “establish confirmation points,” and “develop the necessary competencies” require other actions to be implemented, which are beyond existing planning and control methods. By contrast, the proposed framework has also pointed out the limitations of the role of some project delivery systems, such as IPD and TVD. Although these encourage collaborative work among stakeholders and contribute to the alignment of stakeholders’ interests and goals through governance mechanisms, they do not encompass all principles and prescriptions that are necessary to make planning and control systems effective.

Regarding the theoretical implications, the artefact is based on core production management concepts, originated from Lean Production philosophy (e.g., pull, flow, batch size, and cycle time). The contribution of this investigation is concerned with producing prescriptive knowledge, i.e., design principles and prescriptions, related to the management-as-organizing approach for planning and control systems, which connects those concepts in a meaningful way. This body of knowledge contrasts with the existing literature on the management of ETO industrialized building systems, which presents different types of recommendations in a fragmented way.

Another contribution is related to the fact that some prescriptions can be regarded as innovative approaches for the management of ETO industrialized building systems: (i) pulling the production and delivery of prefabricated components from site assembly as a mechanism to cope with uncertainty; (ii) using the reservation of capacity of manufacturing plants as long-term plans to communicate the demand of site assembly; (iii) emphasizing the role of the monitoring system’s status for establishing confirmation points; (iv) adopting the concept of process modularity in prefabricated building systems to increase efficiency and early detection of problems.

From the perspective of applicability, the framework is relatively easy to understand, as the proposed list of prescriptions is not extensive and is organized into seven categories of general design principles. This framework can be refined and extended, as there may be differences in the prescriptions applicable in different contexts for implementing the same design principles. However, this hierarchical organization should be kept, making it easy to identify chunks of knowledge that are generally applicable from the more specific ones.

6. Conclusions

The main contribution of this investigation is the development of prescriptive knowledge through a set of design principles and prescriptions that can be used to support the design or assessment of planning and control systems that address the requirements of ETO industrialized building systems. This artefact connects several core production management concepts based on the management-as-organizing approach for planning and control systems. Some of the prescriptions can be regarded as innovative for the management of ETO industrialized building systems, including the role of pull production in copying with site assembly variability; how to postpone and trigger actions at manufacturing plants; the role of monitoring the status of production systems; the adoption of the process modularity concept.

Although the adoption of ETO industrialized building systems reduces the complexity of some site assembly tasks and also the variability in some transformation activities carried out in a controlled environment, it introduces other sources of complexity, such as (i) short lead time, which contributes to increase the number of interdependencies between tasks; (ii) uncertainty in relation to client requirements as the order for producing some industrialized components is placed at the design stage; (iii) unexpected conflicts between different building systems and components that need to be installed on-site; (iv) interdependences between different production units, i.e., design teams, manufacturing plants, and site assembly crews. Therefore, the proposed framework represents a planning and control approach aimed at integrating and synchronizing interdependent on-site and off-site processes.

Regarding the limitation of this investigation, it must be pointed out that the artefact is based on a single empirical study, in which there were limitations in the implementation of some of the proposed prescriptions. However, it can be considered a point of departure for other empirical studies that can extend or refine the proposed framework.

Some opportunities for further research emerged from this investigation: (i) test and refine the set of design principles and prescriptions in other empirical studies, including projects that involve a higher degree of industrialization, compared to the context of the project investigated (e.g., modular construction); (ii) explore other planning and control methods and combine them with the ones that have been considered in this investigation so that the set of the proposed prescriptions can be fully implemented; (iii) investigate the systematic use of different types of slack in planning and control systems; (iv) develop approaches that jointly combine the management of both product and operations flows, by adapting practices that have been devised in the manufacturing industry, such as the combination of takt time and standardized work in site installation; (v) devise governance mechanisms (including contracts) that explicitly deal with capacity reservations and confirmation points, in which there are incentives for reliable deliveries.