Task planning and control in construction: revealing workers as early and late planners

Abstract

Production planning and control are critical to project success. They are conventionally understood as linking planning and production by coordinating tasks prior to installation work. However, research has highlighted workers as autonomous and decentralized planners of installation work, which challenges conventional management perspectives. Performing task-level planning and control requires resources and effort, reducing the time spent on installation work. This study explores workers’ planning practices by examining the use of higher-level plans and other information sources and by analyzing workers’ information needs and the factors determining them. A survey of Finnish construction workers was conducted. The results show that some workers prefer planning tasks before arriving at the construction site, while others prefer planning tasks during on-site work preparation. Trade- and crew-specific factors influence planning earlier or later and determine the extent to which centralized and decentralized sources and practices are used. Decentralized sources and practices are most often used to meet workers’ information needs, although centralized sources and practices are also used. These findings may help construction stakeholders better understand the information sources and practices used by workers. Thus, task planning and control activities can better meet workers’ information needs, allowing for more time for installation work.

Keywords:

• Task planning and control
• lean construction
• information needs
• production planning and control
• decentralization
• survey
• worker

Introduction

Production planning and control (PP&C) are critical for project success. Research related to PP&C has mostly taken a management perspective. However, Lehtovaara et al. (2022) found that workers actively participate in the planning of installation work through autonomous and decentralized activities. These results challenge the central assumption of management-as-planning approaches, in which planning performed by management is considered the last planning stage, coupling planning and production by coordinating tasks and removing obstacles before installation work starts (Johnston and Brennan 1996). Planning performed by workers at the task level has been neglected.

Task-level activities include planning and control actions that support installation work (Josephson and Bjorkman 2013; Kalsaas et al. 2014; Seppänen and Görsch 2022), which we refer to as task planning and control (TP&C) activities. Workers autonomously coordinate these activities in a decentralized manner, including reading and interpreting plans, inspecting and preparing installation areas, coordinating materials and tools, discussing work requirements, and making ready other preconditions that may not be in place yet. These activities are carried out before and during installation work (Görsch et al. 2022). Information required for starting installation work is often not evident from project plans (Hamzeh et al. 2015) or is perceived by workers to be outdated (Loosemoore 2014).

From a lean perspective, project plans are planned hierarchically so that the highest-level plans originate from more centralized structures. At this level, the master plan is created, which includes milestones, activities, and project duration to provide strategic direction for the project (Hamzeh et al. 2008). Its primary purpose is to coordinate the long-term scope between the client and the main contractor and to act as a contractually binding agreement between them (Winch and Kelsey 2005). Phase planning, the next hierarchical planning step, breaks down the project into distinct phases (Hamzeh et al. 2008). Within the lean approach, collaborative planning together with trade contractors can be used to optimize handoffs between trade partners to ensure the achievement of project milestones (Ballard 2000). Such a hierarchical planning approach, incorporating different planning levels, addresses project complexity by dividing it into subproblems (Abou-Ibrahim et al. 2019). Consequently, different planning levels differ in the scope and time horizon of decision-making (Laufer et al. 1994). They involve different planning cycles, modeling assumptions, and levels of detail that are reflected in contractual requirements. A major challenge of this planning approach is maintaining coherence between different planning and decision-making levels, which are usually unlinked and unbalanced in construction (Ballard and Howell 1998).

Lean construction advocates recognize that installation work may not be executed as envisioned in project plans due to unprepared preconditions (Bertelsen et al. 2006; Koskela 2000) and upstream variability (Koskela 1999). Therefore, a more detailed planning approach, for example, in the form of the Last Planner System (LPS) (Ballard 2000), aims to translate long-term objectives into production-ready tasks. By comparing the planned situation with project progress, upcoming tasks can be adjusted to compensate for unrealistic or outdated objectives. Lookahead planning connects master and phase planning with weekly planning. Its main function is to break down objectives from project plans to tasks within a time horizon of two to six weeks prior to task execution (Ballard 2000). It designs operations, identifies constraints, and prepares tasks by establishing preconditions and removing relevant constraints (Hamzeh et al. 2008) to enable the most detailed level in LPS weekly planning. Weekly planning seeks coordinated commitments between management and frontline staff to accomplish planned operations within a period of up to one week (Hamzeh et al. 2015).

Despite these planning details, Lehtovaara et al. (2022) found that even in projects utilizing LPS, workers perceive themselves as planners since they participate in TP&C activities, although constraints should be removed and preconditions provided before commitments are made to start installation work. Accordingly, in practice, the extent to which project planning and PP&C support workers’ information needs to perform installation work remains unclear, as workers individually carry out TP&C activities themselves. This suggests that insufficient consideration is given to workers’ information needs to support TP&C.

This study attempts to fill this research gap by examining the extent to which project and production plans are used and the extent to which workers’ information needs are met. Furthermore, it examines what other information sources and planning and control practices are used to support workers’ TP&C. Overall, the study aims to understand how workers plan in practice by investigating the usage of upper-level plans and TP&C activities at supervisors’ and installers’ levels and analyzing the underlying structures of construction workers’ information needs and the factors determining them. The aim is pursued by answering the following research question: “What information sources and planning and control practices support workers in their installation work?”

The research is divided into three parts. First, a literature study was carried out focusing on the hierarchical planning approach (project planning and PP&C) and the current understanding of TP&C practices. Second, a survey was developed addressing PP&C practices and related information needs among 148 mechanical, electrical, and plumbing (MEP) professionals. The focus on MEP work was chosen because building systems and interior finishing work is considered the most challenging work to coordinate, as it is usually carried out by several independent trade partners with crews operating as independent units (Priven and Sacks 2015). Based on the collected data, points of interest were statistically analyzed, interpreted, and discussed. Finally, conclusions are drawn, and suggestions for future research are offered.

Literature review

Project planning

Planning is critical for project success. The tasks of various actors are coordinated before execution to meet project objectives. Planning includes establishing project milestones, activities, and their dependencies and calculating project duration (Hamzeh et al. 2008). Project plans and schedules are part of a contractually binding agreement between the client and the contractor that provides strategic direction by coordinating the project’s scope (Winch and Kelsey 2005). Construction schedules are one of the outcomes of project planning. They describe what tasks (activities), how (methods), when (sequence and schedule), and by whom (resources and competence) will be performed prior to task execution (Alves et al. 2020).

The master plan is the first result of project planning and is created during the pre-construction phase, when planning decisions are made at the macro level. Typically, owners and contractors participate in this planning phase (Hamzeh et al. 2008; Johansen and Wilson 2006). A contract between the client and the contractor specifies the scope and details of the master plan, which can vary widely. The conceptual coordination function of the master plan primarily results in a level of detail and a time horizon in the plans that are sufficient for owner-contractor coordination, but not necessarily for the purpose of day-to-day construction activities (Watkins et al.2009).

Thus, master schedule-level planning can be executed by office-based construction planners with limited or no direct contact or experience with field operations (Johansen and Wilson 2006). Johansen and Wilson (2006) reported that schedulers rely primarily on their personal experiences from previous projects. Collaborative planning is also highly valued but mainly limited to management levels rather than to trade partner levels (Johansen and Wilson 2006; Lehtovaara et al. 2022). Alves et al. (2020) reported that project planners aim to create a traceable project baseline that can be used as a contractual tool to hold people accountable. Therefore, from the client’s perspective, the schedule is less evaluated for accuracy or achievability in terms of production (Johansen and Wilson 2006) since it is used more for coordinating the activities of owner and contractor and to serve as the contractual starting point for delay and change management.

Ballard (2000) described the main goal of project planning as establishing project feasibility throughout the project’s timespan. Its purpose is not necessarily to plan production in the finest detail. Laufer and Tucker (1987) pointed out that very detailed planning well before execution creates additional production and control costs because it is difficult to maintain a clear picture of the project and is usually pointless because of unquantifiable uncertainties. Initial planning should be done at the lowest possible level of detail (Johansen and Wilson 2006).

Each stage of planning fulfils varying purposes by addressing the needs of different users (Laufer et al. 1994). Accordingly, project plans have a different level of detail than production plans. The master plan is used as a project management tool to ensure the client that the plan is realistic and that there are no undue risks of project time or cost overruns. For the contractor, the plans are used in the bidding phase to estimate project costs. Its main goal at this stage is to generate revenue by winning and successfully completing profitable contracts for later phases of the project (Johansen and Wilson 2006). Production plans, instead, require more detail to ensure efficient execution. These details cannot be known during the initial planning phase because they depend on dynamically changing circumstances.

However, the prevailing approach to construction management focuses heavily on the creation, revision, and implementation of plans, commonly referred to as management-as-planning (Johnston and Brennan 1996). This approach assumes that everything can be successfully executed as planned. This assumption has been heavily criticized in the literature (Ballard and Howell 1998; Johnston and Brennan 1996) because it is difficult to fully anticipate or account for unforeseen circumstances. Uncertainty and necessary measures to minimize or mitigate its effects are often neglected (Ballard et al. 2008)

Lean construction theory acknowledges uncertainty in high-level planning and advocates detailing and translating planned objectives into production operations closer to their execution (Alarcón 1997). Lean construction practices, in the form of the LPS, introduce another level of project planning, phase planning, which mediates between long-term and short-term planning so that the project perspective is translated into a production perspective (Hamzeh et al. 2008).

Production planning and control

The LPS production system recognizes that further planning and control are needed to increase planning reliability, boost production performance, and ensure consistent and smooth workflow while facing uncertainty in design and construction operations (Ballard 2000; Hamzeh et al. 2008). Thus, the planning process becomes more detailed and decentralized through collaborative planning, incorporating two more planning levels: lookahead planning and weekly (commitment) planning.

Lookahead planning establishes a link between the phase plan objectives and their implementation through the required activities and preconditions at the operational level in weekly plans. By comparing the planned situation with project progress, upcoming tasks can be adjusted to compensate for delays, outdated objectives, etc. (Olivieri et al. 2019). At this level of planning, operations are designed, constraints are identified, and tasks are prepared by establishing preconditions (Bertelsen et al. 2006; Koskela 2000) and removing relevant constraints (Hamzeh et al. 2008) within a time horizon of two to six weeks so that a workable backlog of implementation-ready tasks becomes available for weekly planning (Ballard 2000).

At the most detailed level of LPS planning, called weekly or commitment planning, the goal is to get commitment from the last planners, who are often crew supervisors, to assignments included in the weekly plan (Aslesen and Tommelein 2016). This includes verifying and ensuring the reliability and quality of made-ready tasks through collaborative efforts between the management and crew levels (Ballard and Howell 1998; Hamzeh et al. 2008). It aims to create a commitment from frontline staff that planned tasks are feasible and can be completed within a period of up to one week (Hamzeh et al. 2015).

LPS assumes that tasks should only be assigned if they can be implemented and are free of constraints, and preconditions for starting installation work are in place so that no further planning effort is necessary. Nevertheless, Lehtovaara et al. (2022) found that workers, even in projects where LPS is used, perceive the need to engage in additional planning and control activities.

Task planning and control

The most detailed level of planning in LPS (commitment planning) does not seem to be sufficient to carry out installation work without delays or interference from other activities (Grau et al. 2020). Instead, workers spend a considerable amount of time on TP&C activities to start and facilitate installation work (Görsch et al. 2022). Several studies have reported a gap between managers’ and workers’ understanding of the information needed for task performance (Caldas and Soibelman 2002; Gil et al. 2000; Tsao et al. 2004). The managerial perspective is often investigated, seeking to provide universal and product-oriented information derived from a more hierarchical, formal, centralized planning process. This information alone does not always add value to the workers’ installation process (Phelps 2012), which may be a reason for decentralized TP&C activities.

Loosemore (2014) reported that workers find the given work plans insufficiently updated and sometimes irrelevant for installation work. Weekly work plans are more concerned with the interdisciplinary coordination of different trade partners than with the installation process of frontline workers on a daily or even more detailed basis (Grau et al. 2020). Planning and control in such detailed and short cycles require a high degree of problem understanding and technology (Grau et al. 2020), as it seems impractical for site managers and engineers to spend a great deal of time tracking such small work packages in detail (Tang et al. 2014). It is currently inefficient and inconvenient to capture data and analyze installation processes during construction at the level of granularity required to accurately monitor installation processes (Sacks et al. 2010). Thus, the causes of incomplete tasks are often not captured and analyzed because of pressure and the resulting time constraints. Consequently, weekly progress reports cannot adequately capture detailed changes in installation work and respond proactively to them (Tang et al. 2014).

Due to continuous dynamic changes, work crews autonomously manage TP&C activities to start or continue installation work. This can encourage improvisation (Hamzeh et al. 2019), which can have a positive impact on installation work but can also lead to waste due to missing preconditions (Formoso et al. 2017). Görsch et al. (2022) analyzed the workday in plumbing work and found that installation work was interrupted, on average, every 3.5 minutes due to insufficient site conditions and preconditions. This was compensated for by additional TP&C activities to continue installation work. Kalsaas et al. (2014) observed how much time workers from different trades spent on different activities, reporting an average of 10–25% on TP&C activities, depending on whether logistical coordination tasks were considered TP&C activities. Josephson and Björkmann’s (2013) study on plumbing work found that about 15–45% of work time was spent on TP&C activities. Several studies have found that additional and unnecessary TP&C activities due to unmet information needs and inadequate preconditions cause increased stress and pressure among workers, which may also be associated with low productivity (Harstad et al. 2015; Loosemoore 2014; Phelps 2012).

Workers are aware that project and PP&C plans are primarily tailored to other perspectives rather than to their own information needs. Nevertheless, workers need the right and relevant information at the right time in relation to task requirements and project progress to enable constraint-free installation work (Gluch and Raisanen 2009; Tenah 1986). Workers respond to unmet information needs by adopting self-guided information gathering and processing practices, such as on-site collaboration and communication (Kania et al. 2020), improvisation (Formoso et al. 2017; Hamzeh et al. 2019), and autonomous decision-making (Lehtovaara et al. 2022). Such practices operate on an even more decentralized level than at the PP&C level to incorporate the information in an ad hoc manner from as many sources as needed to perform the required installation work (Ben-Alon et al. 2014). Furthermore, they are characterized by a low level of advance planning, a go-with-flow attitude, and tight and fast communication structures that lead to opportunistic behavior and allow ad hoc decisions in the field (Ben-Alon and Sacks 2017). According to Lehtovaara et al. (2022), installers perceive site managers less as decision-makers and more as information hubs and work facilitators, leading to perceptions of low hierarchical, informal, and decentralized structures for information gathering and decision-making at the task level.

However, relying on the perceptions and decision-making of autonomous individuals can also have drawbacks. Assigning decision-making authority to a multitude of actors can lead to increased complexity and coordination efforts (Lanaj et al. 2013). Furthermore, inconsistencies in progress tracking (Barber et al. 1999), impeded information flow, and knowledge sharing (Mintzberg 1983) can result from the inclusion of a variety of subjective perspectives, which can lead to a silo mentality and communication problems between disciplines (Lehtovaara et al. 2022). The disadvantages occur mainly when there are no appropriate structures and incentives for decentralized structures (Salovaara and Bathurst 2018) Therefore, it is important to understand the decentralized task-level structures and situational information needs of workers, as this has a significant impact on the smooth workflow, the level of wasted effort, and overall productivity (Dave et al. 2014; Harstad et al. 2015). In what follows, decentralized structures are defined as TP&C activities that are autonomously performed by a large number of workers. By contrast, more centralized practices and sources originate from the hierarchical planning process, project, and PP&C levels. These tend to involve fewer planning authorities than TP&C and stem from more formal, predefined planning and control structures. Figure 1 synthesizes the above-described understanding of the different hierarchical planning levels.

Figure 1. Synthesis of hierarchical planning levels and TP&C.

A diagram describing the different planning levels and their associated plans: Project Planning (master and phase schedule), PP&C (lookahead and weekly planning) and TP&C (continuous planning) depending on the level of detail, time horizon, collaboration and centralization.

Despite these research findings and current understanding, construction management still predominantly manages operational task management and workers’ information needs based on the artifacts of the higher hierarchical and centralized project and PP&C planning levels. The impact of inadequate plans on task performance has been reported (Görsch et al. 2022; Harstad et al. 2015; Josephson and Bjorkman 2013; Kalsaas et al. 2014; Lehtovaara et al. 2022; Loosemoore 2014). However, there is little research on what information sources and planning and control practices help workers coordinate their assembly work. The extent to which project and PP&C plans, or other forms, can be used by workers remains unclear. Further, evidence of construction management meeting workers’ information needs and the structures that underlie those needs is scant. Since workers address their needs through self-directed information gathering, this study also focuses on how they accomplish this.

Methodology

Previous studies have chosen different ethnographic and technological approaches to investigate information sources and planning and control practices from various perspectives. Phelps (2012) studied behavioral factors influencing information needs from a qualitative perspective through observations and interviews. Others have focused on improving the understanding of planning and control patterns and information needs by interviewing managers and supervisors with a scope of either single- or multi-case studies (Ben-Alon and Sacks 2017; Gluch and Raisanen 2009; Harstad et al. 2015). Various social network analyses have analyzed communication and network structures from a cross-hierarchical perspective in individual design and production projects (Flores and Ludlow 1980; Hickethier et al. 2013; Kania et al. 2020; Lehtovaara et al. 2022). Several surveys have also quantitatively analyzed information flows, communication patterns, and specific planning and control practices in design and production projects from the perspectives of managers and supervisors (Hamzeh et al. 2019; Tenah 1986).

Nevertheless, workers’ individual TP&C practices and the factors underlying their information needs and planning and control practices have been less studied from an industry-wide perspective and at the level of workers directly affected. Therefore, an online survey approach was chosen to attract participants exclusively from the installer and supervisor levels and from many different construction projects across Finland.

Survey

The survey was designed following a six-step research design process proposed by Forza (2002), which was extended by two steps integrating an ethical review and a multilanguage approach. The eight steps are (1) literature review, (2) survey design and validation, (3) ethical review, (4) testing, (5) translation and cross-cultural adaptation, (6) data collection, (7) data analysis, and (8) reporting. The reviewed literature led to the formulation of the central research question. This led to the next step of generating the first survey draft based on different survey guidelines (Singleton Jr. and Straits 2018). The survey was then validated in workshops with an industry consortium of 21 companies. A revised draft was submitted to the university’s Research Ethics Committee for an ethical review, leading to the exclusion of personal data-related questions. Afterward, a pilot test was conducted with different test groups—six students, four supervisors, and four installers—whose feedback was used to make adjustments, such as reducing the response time, simplifying language to adapt to the workers’ vocabulary, and setting logical rules for answering the survey.

The survey was conducted in Finland. Due to the many Estonian and other international construction workers in Finland, the survey was conducted in different languages. The survey was originally drafted and revised in English. The final version was translated into Finnish and Estonian and cross-culturally adapted by a team of bilingual speakers with a background in construction, following a standardized procedure (Beaton et al. 2000). The Finnish and Estonian versions were back-translated into English to compare them with the original English version. Based on this comparison, the translation team held discussions and performed iterations to improve the survey.

The target group of the study was defined by adopting a purposeful sampling approach (Patton 2014). Crew heads (supervisors) and crew members (installers) coordinate and perform installation work, which defines them as the target of the study. Due to the COVID-19 pandemic, the survey was conducted online and remained open for data collection for five months, from October 2020 to February 2021. Thus, to reach MEP trades, we chose to distribute the survey through channels of MEP trade unions and employer associations via email links to ensure wide and diverse participation across the Finnish construction industry.

Due to not collecting personal data, distribution via trade unions and employer associations allowed an invitation-based approach to access control within a closed community. Several measures were taken to mitigate potential sampling biases. Distribution through trade unions and employer associations made it possible to invite key subgroups (electrical and plumbing/heating, ventilating, and air-conditioning (HVAC) installers and supervisors) so that they had an equal opportunity to participate in the survey. Further, survey instructions, consent forms, and invitation emails were clearly and transparently worded to clarify the purpose of the survey and the importance of participation by emphasizing that all responses are valuable, regardless of personal opinion or experience, in order to reach a broad range of respondents. Anonymity and confidentiality of responses were communicated and assured to increase willingness to participate and respond truthfully and to reduce potential self-selection bias. To further reduce self-selection bias, the complexity and length of the survey were limited through pilot testing with potential participants and adjusted according to their feedback. Another measure was to send follow-up emails through the various survey distribution channels to remind participants of the importance of their opinions. Finally, demographic data and analyses were used to identify sampling biases that were not found.

The survey instrument included general and specific questions. General questions targeted the demographics of participants, such as role, field of expertise, experience in that field, and type of construction project currently working on. Although the general questions were intended to be answered from the perspective of the participant’s current project, the underlying production system was not addressed for two reasons. First, the survey did not address a specific project. Second, in the testing phase, it became clear that individual construction workers were not often aware of the system management used for planning. Asking details about the PP&C performed by management from workers could have led to inaccurate answers and analyses.

Specific questions targeted utilization rates of schedules, information received and needed during task completion, and the involvement of distinct roles in the flow of information. Closed questions were asked with extra fields for additional open responses (see Appendix). The questions were not compulsory; therefore, certain questions were missing answers. These missing data were analyzed and found to be missing at random, which led to the listwise deletion of non-responses per question. Due to the varying nature of the survey questions and research objectives, different statistical analysis methods were used.

Statistical methods and analysis

Construction schedules are considered one of the main outcomes of the hierarchical planning approach. Construction management predominantly manages operational task management and workers’ information needs based on the schedules of the higher hierarchical and more centralized project and PP&C planning levels. However, workers engage in planning through their own and individual TP&C practices. Thus, the extent to which project and PP&C plans and other forms are used remains unclear. This raises the question of how often workers use schedules, such as master, phase, lookahead, and weekly schedules, as well as other types of planning practices. Thus, a chi-squared test was conducted to assess use differences between project and PP&C plans and other TP&C plans among installers and supervisors.

When little is known about a phenomenon, cluster analysis is often used to provide information about existing associations and patterns in a data sample. It aims to reveal similarities in participants’ behavior and understanding, such as needed types and timing of information for installation work (Jang et al. 2020). Subsequently, in the second step, the survey explored eight types of information needs during five different stages of task performance (see Appendix Question 11), which enabled us to perform a two-step cluster analysis with 40 factors.

Different clustering methods were analyzed before we decided to use a two-step cluster analysis due to its advantages over hierarchical (Balderjahn et al. 2018) and K-means (Sarti et al. 2018) clustering. One advantage is the statistical goodness-of-fit measure in two-step cluster analyses, which determines the number of clusters based on algorithms rather than by subjective decision (Jang et al. 2020). Based on the results of the two-step cluster analysis, we performed Pearson chi-squared tests (cross-tabulations) to further explore workers’ information needs and sources for installation work and how they are reflected in the use of TP&C practices. All tests were conducted using IBM SPSS Statistics version 27.

Demographics

Following previous survey practices (Olivieri et al. 2019), the survey aimed for a minimum of 100 valid responses. The trilingual survey, delivered via an email link distributed by labor unions and employer associations, recorded 1,487 clicks, with 295 respondents starting the survey and 148 completing it. After screening and cleaning (two cases were dropped because they were not part of the target population), a response rate of 9.8% (146 responses) was computed. The response rate was calculated based on the total number of people clicking the link instead of the total number of people exposed to the survey link, which was not possible to track. In Finland, almost all electricians and plumbers belong to a union. The unions report having a total of 41,000 members, which can be used to estimate a national response rate of 0.4%.

Table 1 shows the sociodemographic characteristics of the population sample. Most participants worked on electrical (58.9%) and plumbing (34.9%) tasks, and a smaller number of workers in the field of HVAC (6.2%). Most of the respondents were from the installer level (63.7%), followed by a smaller percentage from the supervisor level (36.3%). A large proportion (68.5%) had a piecework contract, and 31.5% had an hourly contract. Regarding crew sizes, respondents often work in teams bigger than 10 workers (33.6%) or in smaller teams with 2–3 workers (32.2%). Work experience in years is considered an indicator of skill level. Its distribution among the participants was mainly concentrated on medium (4–10 years) and long-term experience levels (11–15 and 16–25 years). Workers with the most experience (over 25 years) were the least represented, as were workers with little experience (0–4 years).

Table 1. Sociodemographic characteristics of respondents (N = 146).

	% (N)
Profession
Electrician	58.9 (86)
Plumbing	34.9 (51)
HVAC	6.2 (9)
Role
Supervisor	36.3 (53)
Installer	63.7 (93)
Experience
0–3 years	10.3 (15)
4–10 years	30.1 (44)
11–15 years	21.2 (31)
16–25 years	24.0 (35)
over 25 years	14.4 (21)
Crew Size
I work alone	15.1 (22)
2–3 workers	32.2 (47)
4–5 workers	11.6 (17)
6–10 workers	7.5 (11)
over 10 workers	33.6 (49)
Construction Project
New Building	58.9 (86)
Renovation	41.1 (60)
Construction Type
Residential Construction	17.1 (25)
Industrial Construction	15.8 (22)
Office Construction	11.0 (16)
Construction of educational institutions	9.6 (14)
Construction of research & laboratory buildings	8.2 (12)
Construction of shopping centers	4.1 (6)
Construction of multi-purpose buildings	15.1 (22)
Infrastructure project	8.2 (12)
Healthcare building	9.6 (14)
I do several projects at the same time	1.4 (2)
Missing Case	0.7 (1)
Working Contract
Piecework Contract	68.5 (100)
Hourly Payment	31.5 (46)
Company Size
I work alone	3.4 (5)
2–5 employees	6.2 (9)
6–15 employees	16.4 (24)
16–50 employees	19.8 (29)
51–100 employees	9.0 (13)
over 100 employees	45.2 (66)

Results

Installers’ and supervisors’ schedule usage

A Pearson chi-squared test (cross-tabulation) was conducted to investigate the differences between supervisors’ and installers’ schedule utilization rates. We also calculated a two-sided Pearson chi-squared significance value (p-value; significance level of 0.05 or higher) per schedule type. A total of 141 responses were included due to 5 incomplete responses. Table 2 shows statistically significant differences between the supervisors’ and installers’ utilization rates for phase, look-ahead, weekly, and self-planned schedules.

Table 2. Schedule utilization frequencies by role; Note: I = Installer; S = Supervisor.

		Very often/Often % (N)	Medium/Low usage % (N)	Occasional/No usage % (N)	p-value
Master schedule	I	19.1 (17)	52.8 (47)	28.1 (25)	0.091
	S	32.7 (17)	51.9 (27)	15.4 (8)
Phase schedule	I	21.3 (19)	50.6 (45)	28.1 (25)	0.009
	S	44.2 (23)	42.3 (22)	13.5 (7)
Look ahead schedule	I	29.2 (26)	43.8 (39)	27.0 (24)	0.005
	S	51.9 (27)	40.4 (21)	7.7 (4)
Weekly schedule	I	27.0 (24)	44.9 (40)	28.1 (25)	0.003
	S	53.8 (28)	34.6 (18)	11.5 (6)
Self-planned schedule	I	39.3 (35)	40.4 (36)	20.2 (18)	0.022
	S	63.5 (33)	25.0 (13)	11.5 (6)
No written schedule/I’ve got it in my mind	I	39.3 (35)	40.4 (36)	20.2 (18)	0.121
	S	45.3 (29)	34.8 (13)	19.9 (10)

Schedules provided through more centralized practices were more often used by supervisors than by installers. The different planning levels were used ‘very often/often’ by a maximum of 53.8% of the supervisors and 29.2% of the installers. Furthermore, Table 2 shows that the frequency of use increased as the level of detail in schedules provided by management increased: Master schedules (lowest level of detail) were not used very often (only 19.1% of installers and 32.7% of supervisors used them “very often/often”), while weekly schedules (highest level of detail) were used more often (27.0% of installers and 53.8% of supervisors). “Self-planned schedules” and “No written schedule/I’ve got it in my mind” are decentralized TP&C practices and can be used in parallel with more centralized provided schedules. They were used “very often/often” most often by installers and supervisors (“Self-planned schedules”: 39.3% installers and 63.5% supervisors “No written schedule/I’ve got it in my mind”: 39.3% installers and 45.3% supervisors). Table 2 further reveals that installers did not record their plans in writing (self-planned schedule share equals no written schedule share), while supervisors more often recorded their plans (larger share of “Self-planned schedules” than “No written Schedule/I’ve got it in my mind”).

These results indicate that while centralized-provided schedules were used at the task level, decentralized TP&C practices coexisted and seemed to support individuals’ installation work. Furthermore, the installers regularly conducted TP&C activities, which are unanticipated in management-as-planning approaches due to the assumption that planning performed by management is the last planning stage. Thus, to support installers’ TP&C activities to potentially shorten or increase their effectiveness, the first and second observations underline the importance of understanding workers’ situational information needs.

Clustering patterns of workers’ information needs

Aiming to find similarities in participants’ TP&C behavior, we conducted a two-step cluster analysis using eight types of information needs (information about task requirements, materials, tools and equipment, work location, occupation status of work location, time until work location change, task duration, and available workforce) during five different stages of TP&C performance (before arrival on site; during arrival on site; during task preparation; during task execution; after task is finished; see Appendix Question 11). The SPSS automated two-step cluster analysis revealed two clusters. This was manually confirmed post hoc by testing solutions with 1–10 cluster structures and comparing the individual silhouette values of each solution. The two-cluster solution showed the highest silhouette value of 0.3. The silhouette measure of cluster cohesion and separation of 0.3 demonstrates fair evidence of cluster formation (the silhouette measure of cohesion and separation needs to be greater than 0.0 for intra-cluster spacing and inter-cluster spacing to be valid) (Norusis 2011). The analysis showed equally sized clusters: Cluster 1 contained 50.7% (74 participants), and Cluster 2 contained 49.3% (72 participants), with a ratio size of 1.03. The factors with the highest predictor importance related to the cluster were information needs during task preparation and before arrival on site. Factors with medium importance were information during task execution and handover, followed by information needed during arrival on site with the lowest importance.

The clusters differentiated from each other in terms of respondents’ information needs before arriving on site and when preparing tasks. The information needs of Cluster 1 peak before arriving on site. Information about the work location (91.7%), task requirements (86.1%), and available workforce (87.5%) are demanded the most, followed by information about the task duration (80.6%) and location changes (79.2%). Cluster 2’s information needs peaked later, during task preparation. Information about materials (90.5%) and task requirements (89.2%) was demanded the most, followed by information about the available workforce (86.5%) and task duration (85.1%). In practice, this means one cluster requires information before arriving on site, and the other cluster while preparing tasks. Based on the timing of their information needs, the clusters were named early planners (Cluster 1) and late planners (Cluster 2).

Table 3 shows the differences between the clusters based on Pearson’s chi-squared test. In the population studied, early planners (n = 74) were more often found in small crews (p = 0.010) or individual workers (company size: I work alone = 80.0%). Most of the plumbing (62.7%) and HVAC (77.8%) workers were statistically significantly identified as early planners (p = 0.005). Other notable characterizations of early planners, but not statistically significant, were their overrepresentation among hourly-paid workers (58.7%) and supervisors (56.6%). The second cluster, late planners (n = 72), had a high proportion of electricians (statistically significant). Larger crews, with more than 10 workers, statistically significantly more often included late planners (69.4%) than early planners. Late planners were more often found in larger companies (51–100 employees: 61.5%; over 100 employees: 57.6%). No differentiation between the clusters was found in their average project and progress understanding, working hours, construction projects, and workers’ experience. Table 3 summarizes all the results described above.

Table 3. Pearson chi-squared test of early and late planners.

	Early planner, % (N)	Late planner, % (N)	p-value
Profession^Table Footnotea			p = 0.005
Electrician	38.4 (33)	61.6 (53)
Plumbing	62.7 (32)	37.3 (19)
HVAC	77.8 (7)	22.2 (2)
Role			p = 0.184
Supervisor	56.6 (30)	43.4 (23)
Installer	45.2 (42)	54.8 (51)
Experience			p = 0.783
0–3 years	46.7 (7)	53.3 (8)
4–10 years	56.8 (25)	43.2 (19)
11–15 years	41.9 (13)	58.1 (18)
16–25 years	48.6 (17)	51.4 (18)
over 25 years	47.6 (10)	52.4 (11)
Crew Size			p = 0.010
I work alone	72.7 (16)	27.3 (6)
2–3 workers	51.1 (24)	48.9 (23)
4–5 workers	58.8 (10)	41.2 (7)
6–10 workers	63.6 (7)	36.4 (4)
over 10 workers	30.6 (15)	69.4 (34)
Construction Project			p = 0.417
New Building	46.5 (40)	53.5 (46)
Renovation	53.3 (32)	46.7 (28)
Construction Type^Table Footnotea			p = 0.170
Residential Construction	52.0 (13)	48.0 (12)
Industrial Construction	31.8 (7)	68.2 (15)
Office Construction	56.3 (9)	43.7 (7)
Educational building	64.3 (9)	35.7 (5)
Research & laboratory buildings	33.3 (4)	66.7 (8)
Shopping centers	66.7 (4)	33.3 (2)
Multi–purpose buildings	63.6 (14)	36.4 (8)
Infrastructure project	18.2 (2)	81.8 (9)
Healthcare building	53.3 (8)	46.7 (7)
I do several projects at the same time	50.0 (1)	50.0 (1)
Missing Value	0.0 (0)	100.0 (1)
Working Contract			p = 0.155
Piecework contract	45.0 (45)	55.0 (55)
Hourly Payment	58.7 (27)	41.3 (19)
Company Size^Table Footnotea			p = 0.248
I work alone	80.0 (4)	20.0 (1)
2–5 employees	44.4 (4)	55.6 (5)
6–15 employees	66.7 (16)	33.3 (8)
16–50 employees	51.7 (15)	48.3 (14)
51–100 employees	38.5 (5)	61.5 (8)
over 100 employees	42.4 (28)	57.6 (38)
Working Hours^Table Footnotea			p = 0.973
1–4 hours	60.0 (3)	40.0 (2)
5–6 hours	42.9 (3)	57.1 (4)
7–8 hours	50.0 (56)	50.0 (56)
9–10 hours	44.4 (8)	55.6 (10)
over 10 hours	50.0 (2)	50.0 (2)
Average Progress Understanding^Table Footnotea			p = 0.897
Very Poor	100.0 (1)	0.0 (0)
Poor	58.3 (7)	41.7 (5)
Average	50.0 (28)	50.0 (28)
Great	46.6 (27)	53.4 (31)
Excellent	47.4 (9)	52.6 (10)

^aFisher’s exact test, due to lower count cell than five; p-values below 0.05 represent statistically significant findings.

On average, 90.1% of early planners identified all eight types of information as needed before arriving on site, and 20.0% also as needed during task preparation. By contrast, an average of 35.5% of late planners identified all eight types of information as necessary before arriving on site, and 77.8% also as necessary during task preparation. The information needs of early planners are more focused on one particular stage, while late planners show a somewhat lower concentration of information needs in one particular stage.

Early and late planners’ schedule usage

Demanding information early is correlated with higher usage of more centralized planned schedules, based on results comparing the utilization frequencies of the schedules of early and late planners. Table 4 shows that early planners have a higher share of installers who use all schedule categories more often “very often/often” than late planners. Furthermore, both early and late planners use decentralized TP&C practices more often than more centralized schedules in the category “very often/often” (early planners slightly more planned than late planners). Statistically significantly, more late planners use self-planned schedules with a “medium/occasional” frequency, and earlier planners use self-planned schedules with a “low/no usage” frequency.

Table 4. Schedule usage of early (EP) and late planners (LP).

		Very often/Often	Medium/Occasional	Low usage/No usage	p-value
Master Schedule	EP	.282	.437	.282	.148
	LP	.222	.597	.181
Phase Schedule	EP	.366	.380	.254	.107
	LP	.250	.556	.194
Look ahead Schedule	EP	.423	.338	.239	.127
	LP	.347	.500	.153
Weekly Schedule	EP	.394	.366	.239	.634
	LP	.347	.444	.208
Self-planned Schedule	EP	.507	.239	.254	.008
	LP	.458	.444	.097
No written Schedule, I’ve got it in my mind	EP	.465	.324	.211	.906
	LP	.451	.366	.197

These results indicate that early planners tended to more frequently utilize information originating from centralized practices. Late planners tended to more often use information created through autonomous and decentralized TP&C practices in the form of “self-planned schedules”. “No written Schedule/I’ve got it in my mind” seemed to be a common and equally often used practice among both early and late planners. Overall, both groups used both centrally planned schedules and decentralized TP&C practices, but decentralized TP&C practices were used more often than more centralized planned schedules.

Early and late planners’ perceived information distribution

For an in-depth understanding of how early and late planners differed in their TP&C practices, the workers’ perceptions of wanted and received information per stage of task performance were analyzed. The workers’ perceptions of received information from Question 10 (“What information do you receive in each of the following situations”) were compared with their perceptions of wanted information from Question 11 (“What information do you need in each of the following situations”). This resulted in cases of matching (“wanted and received” or “not wanted and not received”) and unmatching (“wanted, but not received” or “received, but not wanted”) information supply and demand per information and task completion stage. Focusing on the unmatching perspective, Figure 2 shows the share of early planners who perceived an information need without receiving it and their share of receiving information without a need throughout five different task performance stages. Figure 3 shows the same for late planners.

Figure 2. Early planners’ perceptions of wanted and received information.

A graph showing early planners’ wanted, but not received and received, but not wanted types of information in percentage during five different task performance stages.

Figure 3. Late planners’ perceptions of wanted and received information.

A graph showing late planners’ wanted, but not received and received, but not wanted types of information in percentage during five different task performance stages.

Early planners were not well served by the information supplied, and information was typically unavailable before arrival on site. The information most often perceived as wanted but not received was workspace-related information, such as the occupancy status (55.6%) and change requirements (50.0%) of the work location, available workforce (54.2%), task length (52.8%), and requirements (45.8%).

Another trend shown in Figure 2 is higher shares of received information without being wanted throughout all later task performance stages. Especially in the “preparation” phase, these proportions were apparent when compared to “before arrival”. Information about task requirements (34.7%), tools and equipment (29.2%), and materials (27.8%) was received too late, since they were wanted earlier. Few early planners needed information during “execution” and “handover”, yet many received information that late.

Late planners differed from early planners in needing more information at later stages (during arrival and preparation), while early planners needed the most information in the first stage. Both received information too late. The receipt of very-late information by late planners occurred mostly during the preparation stage. Here, workers needed work location-related information, such as occupancy status (36.5%) and change requirements (37.8%) of the work location, available workforce (33.8%), and task duration (35.1%). During the “execution” and “handover” phases, the proportion of information wanted but not received was lower in comparison to earlier phases, while the proportion of information received but not wanted rose sharply with time, indicating that information was received too late. Here, work location-related information like task duration (37.8%), occupancy status (31.1%), and change requirements (28.4%) of the work location, and task requirements (27.0%) are perceived as the most delayed types of information.

In summary, Figure 2 and Figure 3 show that up to 61.1% of early planners and up to 47.3% of late planners report needed and received information as unmatching in certain instances. The results indicated that unwanted information was supplied to workers, and needed information was provided too late. An information distribution system with such gaps shows the potential for improvement and the necessity for a deeper understanding of the needed information. Therefore, it seems useful to understand how early and late planners acquire information by analyzing the content in communication patterns between workers and other roles.

Topic and role-specific scope of workers’ communication

Table 5 shows the shares of respondents who reported communicating with distinct roles about specific topics. For each combination of role and topic, Pearson’s chi-squared test was conducted to analyze the differences between early and late planners. Statistically significant findings are shown with gray boxes in Table 5. Workspace management (53% on average) and time and task management (48% on average) were the topics most frequently communicated across all roles. Table 5 also indicates that more late planners discussed these topics on average and on crew levels than early planners. In general, late planners were more likely to be communicators, expressed through higher levels of communication. This finding fits well with the notion that late planners’ task preparation relies less on more centralized planned schedules and interacts with more agile decentralized TP&C decision-making via communication at the crew level. By contrast, early planners operate more often in smaller crews or alone, leading to a smaller number of colleagues in their work environments. This could explain why early planners rely on communicating these topics more often with high- and mid-level management.

Table 5. Shares of communication with different roles about certain topics.

	Site CM		CM Main Contractor		Own PM		Own CM		Supervisor		Different Trade		Same Trade		Average
	EP	LP	EP	LP	EP	LP	EP	LP	EP	LP	EP	LP	EP	LP	EP	LP	Total
Workspace Management	0.49	0.53	0.33	0.38	0.54	0.53	0.63	0.62	0.64	0.73	0.44	0.57	0.49	0.66	0.44	0.50	0.53
Time & Task Management	0.54	0.43	0.32	0.30	0.51	0.61	0.63	0.64	0.58	0.64	0.32	0.38	0.40	0.62	0.41	0.45	0.48
Materials	0.10	0.07	0.04	0.09	0.24	0.30	0.40	0.50	0.56	0.68	0.11	0.19	0.47	0.58	0.24	0.30	0.29
Tools and equipment	0.06	0.03	0.07	0.03	0.24	0.23	0.31	0.30	0.51	0.66	0.13	0.18	0.53	0.64	0.23	0.26	0.25
Reporting	0.15	0.15	0.19	0.20	0.32	0.27	0.35	0.36	0.25	0.42	0.04	0.07	0.15	0.28	0.18	0.22	0.23
Safety	0.24	0.26	0.15	0.19	0.21	0.14	0.22	0.22	0.22	0.38	0.14	0.14	0.21	0.38	0.17	0.21	0.21
Chatting	0.06	0.07	0.06	0.05	0.04	0.07	0.08	0.18	0.38	0.46	0.24	0.32	0.51	0.64	0.17	0.22	0.19
Work-related Discussions	0.04	0.03	0.06	0.05	0.04	0.05	0.13	0.11	0.35	0.35	0.32	0.28	0.53	0.54	0.18	0.18	0.18
Quality Control of Work	0.07	0.12	0.14	0.09	0.26	0.11	0.32	0.23	0.24	0.45	0.10	0.07	0.15	0.32	0.16	0.17	0.18

Notes: CM = Construction Management, PM = Project Management.

Considering all topics, both clusters had more communication with task-level labor than with management. This is in line with the results shown earlier, which indicate that more centralized planned schedules were less often used for workers’ task performance than self-prepared schedules supported by on-site communication. Against this background, both clusters favor autonomous and decentralized communication with task-level employees.

Table 5 shows high shares of workers communicating on-site. Communicated topics include safety, tools and equipment, materials, tasks, workspace allocation, and quality control of work. These topics show evidence that workers carry out TP&C activities, including process coordination and progress reporting. In contrast to the assumptions of the management-as-planning approach, workers act as the final planning authority, linking and detailing planning and production by coordinating tasks and removing obstacles before assembly work begins. Workers carry out planning activities in addition to installation work for which they are hired.

Discussion

The following section aims to answer the overall research question, “What information sources and planning and control practices support workers in coordinating their installation work?” The section addresses the central research question in parts by focusing first on the extent to which project and PP&C plans, or other forms, are used by workers. Next, it considers the extent to which construction management satisfies workers’ information needs and what structures underlie those needs. The next step addresses how workers accomplish their information needs through self-directed information gathering. Finally, the limitations of the study and future research are pointed out.

Thus, the study investigated information needs as well as applied planning and control practices supporting installation work. The results show that information acquisition and provision at the supervisor and installer levels are based on both centralized and decentralized information sources and planning and control practices, but decentralized TP&C practices are more often used to meet information needs at the operational level. Based on workers’ information needs, on-site professionals can be clustered into early and late planners, which is the fundamental finding of this paper. We also found that the underlying information needs and clusters were procedural (industry-specific) and organizational (team size) aspects. Electricians, identified mostly as late planners, operate in larger crews that use more frequently decentralized information sources and planning and control practices. Plumber/HVAC workers, mainly identified as earlier planners, operate more often in smaller crews, which tend to use centralized information sources and planning and control practices more frequently than late planners.

Integration of centralized and decentralized information sources

Loosemoore (2014) reported that more centralized information is less often used at operational levels since it primarily provides generic task- and process-relevant information, which is perceived as vague and static for the purpose of TP&C activities (Watkins et al. 2009). The present study confirms these findings by providing evidence that centralized schedules are used less often among installers. Supervisors use centralized information more often because it may be more relevant to their task scope; it incorporates a different time horizon, degree of complexity, and level of detail than installers’ scope. More detailed plans are used more often. This increased usage confirms the purpose of the LPS methodology to shield production and ensure the reliability of plans (Ballard and Howell 1998; Ballard 2000; Hamzeh et al. 2008; Hamzeh et al. 2015; Laufer and Tucker 1987). Thus, this finding shows that providing centralized information refined through the hierarchical planning process adds value to the installer’s TP&C activities and installation work, as they are used often and cover certain information needs to a high degree.

However, the provision of information from only one source, in this case, centralized plans, may not necessarily provide all preconditions for installation work (also noted by Görsch et al. 2022; Grau et al. 2020), which is why it is supported by decentralized sources and practices. Decentralized gathered information is used more often by supervisors and installers than by information originating from centralized practices. These results confirm the findings of Lehtovaara et al. (2022), who showed that supervisors and installers consider centralized and decentralized information sources in parallel. Moreover, the results suggest that in current production systems in construction, information distribution through centralized and decentralized practices intertwines and complements each other.

Furthermore, the high use of self-generated plans, often assumed in the literature (Ben-Alon et al. 2014), shows that installers regularly plan their own actions. This finding is largely unanticipated in management-as-planning approaches, which assume that management provides all the conditions needed to carry out work (Johnston and Brennan 1996). Our findings also reveal installers as the last planning authority. This and the findings of Lehtovaara et al. (2022) challenge the understanding of the role of the last planner. The last planner is often described as the supervisor who is closest to the work and has the decision-making authority to coordinate reliable commitments for task completion (Ballard 2000; Aslesen and Tommelein 2016).

The results also show that supervisors and installers often rely on their personal skills and experience to incorporate ad hoc planning activities (“No written Schedule/I’ve got it in my mind”). Here, plans are created spontaneously and without much preparation to supplement the schedules and instructions provided through centralized practices that do not meet the needs of workers. This can be understood as a form of improvisation. The high use of this practice is consistent with previous findings on subcontractors’ need to apply ad hoc decision-making (Loosemoore 2014) and the importance of improvisation in construction by Hamzeh et al. (2019) and Hamzeh et al. (2018). However, it also highlights the difficulties associated with such a practice, as it might lead to the generation of wasted effort (Formoso et al. 2017).

Decentralized practices can respond in dynamic ways, supporting individual situational understanding and mismatched information needs through approaches such as improvisation, self-planned schedules, and communication. Such approaches rely heavily on individual experience, manual observation, verbal communication, and team- and subject-specific understanding. These practices require time and effort (Seppänen and Görsch 2022). They are not considered process steps in management-as-planning approaches, as it is assumed that all information is available when installation work begins (Johnston and Brennan 1996).

Underlying structures of TP&C information needs

Information needs can be categorized into clusters of early and late planners, which differ statistically significantly between their specifics of work processes per discipline (trade) and structural differences in crews (size). Plumbing and HVAC (mostly identified as early planners, with the highest information needs before arrival) work are described as highly dependent on space constraints relative to task requirements, where unknown space and access constraints prior to the work phase become limiting factors for such subcontractors (Gil et al. 2000). Conservative assumptions regarding the selection and sizing of materials and tools are often made if spatial constraints remain unclear before getting on site. In plumbing work, for example, components often consist of many individual parts that are delivered in small batches, requiring lots of work structuring effort (Tsao et al. 2004) depending on the task and space conditions on site (Görsch et al. 2022). By contrast, the information needs of electricians were highest during work preparation, and they were therefore mostly identified as late planners. In line with this, Seppänen and Görsch (2022) described electrical work as less spatially constrained than plumbing work because there are fewer different components and more flexible materials, which contributes to more adaptive and informal planning and information-gathering strategies later in the process. It is easier to improvise with fewer material types and when coordination is not a big problem.

From a practical point of view, early planning can be considered a TP&C practice that is used when the installation process is subject to severe spatial constraints and uncertainties due to high task requirements based on many individual and limited flexible components. Accordingly, planning starts at an early stage to design, coordinate, communicate, and adapt an appropriate installation process, which can be considered a confirmation of the qualitative results of Gil et al. (2000). Late planning can instead be understood as a TP&C practice that is less constrained by spatial restrictions and applies an installation process based on more flexible components or less customization. Accordingly, planning can start at a later stage of task performance.

Most information needed by early planners is work location, workforce, and task requirements, which can be viewed as work structuring requirements related to spatial constraints prior to arrival on site (Tsao et al. 2004). Late planners, however, are most often in need of material, workforce, and task requirements during task preparation. The information needs of late planners focus less on spatial constraints and more on material handling and installation restrictions. Thus, it appears that spatial and task-related information has a higher priority in the work structuring of early planners. By contrast, late planners perceive spatial constraints as less limiting, so they focus more on work structuring and the preparation itself. Here, task planning is more individualized and varied.

Factors shaping the forms of decentralized TP&C practices

A factor that also significantly differentiates between early and late planners is the size of the crew. The difference in the number of employees affects the amount of communication and the way information is obtained. Early planners, more often found in small crews or as single-entity crews, are more likely to communicate with authorities at higher hierarchical levels, as otherwise, they would rely primarily on their own skills, experience, and situational understanding. Communication is more common in larger crews than in smaller crews as a tool for sharing and considering different situational understandings in terms of time, tasks, and workspace management. Larger crews are made up of a variety of opinions (high degree of decentralization), which can lead to a great deal of consideration of different people’s understanding and knowledge when making decisions within a crew. Considering more individual knowledge and situational understanding may be a more flexible approach that incorporates the latest information and conditions from the field.

Furthermore, in larger crews, the range of individual information needs can be wider than in smaller crews, as shown by the greater variance in the information needs of late planners. This variance may be a different situational understanding depending on the individual’s role in the group (job specification). Therefore, we can assume that it is more difficult to serve information needs with higher variance in crews if a uniform approach is adopted for all of them.

The nature of the information most needed but not received (work location occupancy, task duration, location changes, and available workforce) can be described as dynamically evolving on-site information that is often inadequately detailed, monitored, and updated prior to arrival on site. Thus, the early planners in this study showed the greatest lack of such information prior to arrival, as this information was rarely available at this time (up to 61.1%). Late planners showed similarities in unmet needs because these types of information were often unavailable during the preparation phase (up to 47.3%).

When updated information is rarely available when needed prior to arrival at construction sites, centralized sources of information and communication with higher hierarchical levels become the only sources of information available. From this perspective, and given the small and solo crew structures in which internal communications are rare or non-existent, it seems unsurprising that early planners communicate more with higher hierarchical authorities (high and mid-level management) and incorporate more information from more centralized sources than late planners.

From the perspectives of late planners, who focus greatly on on-site task preparation and act in larger teams, missing or inadequately provided information is compensated for more often through decentralized practices, such as communication, improvisation, and self-planned schedules. Such self-management can react dynamically and ad hoc, especially in cases where the centralized information provided is considered too vague and static (Loosemoore 2014) for the purpose of TP&C activities (Watkins et al. 2009). However, a downside of larger teams was noted by Gong et al. (2011), who indicated that crew effectiveness decreases as crew size increases, as contractors use larger-sized crews to compensate for the lack of detailed TP&C. The above-described strategies of late planners seem to be aligned with the findings of Gong et al. (2011), thus necessitating future investigation of the crew effectiveness of early (smaller crews) and late (larger crews) planners.

Limitations and future research

Overall, the study and its results are limited by the survey’s scope, which is based on differences between electrical, plumbing, and HVAC work in the Finnish construction industry. This could be a Finnish peculiarity, but over the years, other research findings have shown similar problems in many other contexts (Caldas and Soibelman 2002; Grau et al. 2020; Harstad et al. 2015; Josephson and Bjorkman 2013; Kalsaas et al. 2014; Lehtovaara et al. 2022; Loosemoore 2014). Future research should repeat the study in other countries to validate the findings. Such a synthesis of studies and results would help obtain a more holistic view of TP&C structures and improve the validity and reliability of the current results. Moreover, the longitudinal effects of such approaches could be considered in future research initiatives.

Furthermore, the cluster analysis tries to consider the very individual situations in the construction industry, where workers are exposed to constantly changing needs and conditions. Such analysis leads to grouping based on the largest differentiating factors, which is why further subgroups within the clusters could exist that do not fully follow the described clustering due to other variables. In addition, since cluster analyses investigate structural differences within existing datasets, an extension or modification of the dataset may lead to significantly different results. These factors and the nature of the data collected can explain the medium–high silhouette value of 0.3 in this population, which at the same time shows the relevance of both clusters.

Due to the design of the survey (online, voluntary, no collection of personal data), several measures were taken to reduce potential sampling and self-selection bias (see section Survey). Despite these measures, it is challenging to fully avoid potential bias. Accordingly, the validity and reliability of the study should be viewed with caution, especially since it is currently focused on a single country and uses a statistical analysis method that can be sensitive to small changes in a dataset (cluster analysis). The replication of the study in other countries and contexts is emphasized.

The frequency of centralized and decentralized information provided was based only on the different planning practices and sources, which, in this form, did not allow an assessment of the extent to which different types of information were collected through centralized and decentralized practices. Future research could explore forms of information incorporation at the crew and individual levels depending on situational conditions and personal factors affecting decision-making based on centralized and decentralized practices. Semi-structured in-depth interviews can support the exploration of installers’ thoughts, experiences, and understandings, which are often hidden in tacit knowledge and require complex questions and considerable inquiry.

Further, information needed and received during different stages of task performance was collected without considering the extent to which such information adds value to or disrupts workflows and leads to waste. How missing preconditions (information needs) affect workflows and contribute to wasted effort in construction at the operational level may be explored through a comparative, multi-case, quantitative analysis. Information and effective workflow management are subject to project, crew, and individual factors, which is why components such as production system, building type, trade type, crew size, and individual experience need to be considered, affecting the degree of wasted effort in construction.

Conclusion

In this study, we investigated factors affecting the usage of centralized and decentralized TP&C practices at the supervisor and worker levels, as well as at operational levels. This study analyzed micro-level and timely situational information and communication patterns, which have been lacking in the literature. The results make novel theoretical and practical contributions to centralized and decentralized information sources and applied task planning and control practices with respect to timely needs for task performance.

Considering the central research question, “What information sources and planning and control practices support workers in coordinating their installation work?”, first, we found that centralized and decentralized practices are regularly used at both levels, with centralized information sources being used more often as granularity increases. However, decentralized information sources and TP&C practices are used most often at both levels. Thus, installers can be represented as autonomous planning authorities, which are mainly unanticipated in management-as-planning practices, trying to enable and improve their task performance by complementing information that was not sufficiently provided by centralized practices when needed.

Second, we showed that situational information needs for TP&C activities and task performance can be clustered into early and late planners. These clusters differ significantly between discipline and crew size specifics. Furthermore, early planners tend to incorporate information more often through centralized practices. Late planners tend to incorporate information more often through decentralized practices. These characteristics of both clusters were found to be the most distinguishing factors between early and late planners.

Finally, the centralized planning approach does not provide all the information when needed, especially in the early stages of task performance. Thus, workers become independent and individual planners and decision-makers of TP&C activities and installation work, which adds complexity in an already highly complex field often described as suffering from chaotic and low productivity levels. Thus, future research and practical developments should pay more attention to the information needs of workers and acknowledge that weekly planning is not the last stage of planning and control. Decentralized TP&C practices are used to compensate for centralized provided information, which is not sufficient for workers’ installation work. This forces workers to spend time on TP&C activities that are not included in the project and PP&C plans, which reduces installation time. It is assumed that these TP&C activities are highly dependent on subjective skills, experience, observation, communication, and understanding, which can lead to unpredictable results in task performance and uncertainty in the project schedule, which lean construction theory and practices aim to reduce through the hierarchical planning approach. However, the approach’s current level of granularity (detail, complexity, decentralization, and short-term horizon) does not allow for better support of TP&C activities and installation work.

Therefore, researchers and practitioners should focus on how to meet the timely needs of crews and installers in several ways: 1) detailing plans toward the needs of supervisors and installers prior to TP&C activities to provide more concrete information early on; 2) developing ways to allow installers to document and report task performance and constraints, which are currently distributed primarily through decentralized practices; 3) developing dynamic monitoring systems that capture field conditions and progress to reduce the amount of time installers spend gathering this information; and 4) providing visualized and personalized process information to crews and installers to improve their situational understanding. These steps could more efficiently support TP&C practices and allow for more time for installation work in the future.

Acknowledgements

The authors would like to thank Lauri Vittala and Juho Saranpää for their assistance during the data collection and Ergo Pikas and Juho Lahtinen for their assistance in translating the survey. We would also like to acknowledge the contributions of the editor and the anonymous reviewers to the paper’s peer review, which allowed it to be improved significantly.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

Due to the nature of the research, due to [ethical/legal/commercial] supporting data is not available.

Additional information

Funding

This study was conducted as part of a survey funded by the Finnish associations of electrical and telecommunication companies STUL, HVAC and plumbing/HVAC technical companies LVI-TU, and electrical employers STTA.

Appendix:

Survey questions

1. What is your age?

   (18–25 years old; 26–35 years old; 36–45 years old; 46–55 years old; over 56 years old)

2. What kind of construction project are you currently working on?

   (Residential; Industrial; Office; Educational building; Research & laboratory building; Shopping centers; Multi-purpose buildings; Infrastructure project; Healthcare building; I do several projects at the same; Other option)

3. What is this construction project?

   (New building; Renovation project)

4. What is the style of your current working contract?

   (Piecework contract; Hourly payment)

5. How many people does your current employer employ?

   (I work alone; 2–5 employees; 6–15 employees; 16–50 employees; 51–100 employees; Over 100 employees)

6. What trade do you belong to in this construction project?

   (Electrician/supervisor; Data network installer; Data network installer/supervisor; Plumber; Plumber/supervisor; HVAC installer; HVAC installer/supervisor; Other option)

7. How long have you worked in your current role?

   (0–3 years; 4–10 years; 11–15 years; 16–25 years; Over 25 years)

8. How many workers are on your team on this construction project?

   (I work alone; 2–3 workers; 4–5 workers; 6–10 workers; Over 10 workers)

9. How many hours do you spend on site per day?

   (1–4; 5—6; 7–8; 9–10; Over 10)

10. What information do you RECEIVE in each of the following situations (Matrix question):

    (Horizontal Answers: Before first arrival on site; During first arrival on site; Prepare your tasks; Work on your task; When task is ready for handover; Vertical Answers: Information about materials; Information about tools and equipment; Information about work location; Information about occupation of work locations; Information about time till change of work; Information about total duration of task; Information about available workforce; Information about task requirements; Other option)

11. What information do you NEED in each of the following situations (Matrix question):

    (Horizontal Answers: Before first arrival on site; During first arrival on site; Prepare your tasks; Work on your task; When task is ready for handover; Vertical Answers: Information about materials; Information about tools and equipment; Information about work location; Information about occupation of work locations; Information about time till change of work; Information about total duration of task; Information about available workforce; Information about task requirements; Other option)

12. Please describe your UNDERSTANDING in each of the following situations, how your work status influences other trade’s work (on a scale from 1 (poor) to 5 (excellent))

    (Before first arrival on site; During first arrival on site; Prepare your tasks; Work on your tasks; When task is ready for handover)

13. With whom do you communicate about which topics? (Matrix question)

    (Horizontal Answers: Time & task management; Workspace management; Finding tools and equipment, Preparing materials; Work-related discussions with colleagues; Chatting to colleagues (not work related); Quality control of my work; Reporting the progress of my work; Safety Management; other option; Vertical Answers: Construction manager in charge of the site; Construction manager of the main contractor; Own project manager; Own construction manager; Foreman; Worker - with different type of work; Worker - with same type of work; Other option)

14. How often do you make use of the following timetables to prepare your work on a weekly basis? (Matrix question)

    (Horizontal Answers: Very often; Often; Medium usage; Occasionally; Low usage; I don’t use; Vertical Answers: Master schedule; Phase schedule; Look ahead schedule; Weekly Schedule; A schedule I prepared myself; No written schedule, I’ve got it in my mind; Other option)