Content area
Purpose
The aim of this paper was to study current practices in FM work order processing to support and improve decision-making. Processing and prioritizing work orders constitute a critical part of facilities and maintenance management practices given the large amount of work orders submitted daily. User-driven approaches (UDAs) are currently more prevalent for processing and prioritizing work orders but have challenges including inconsistency and subjectivity. Data-driven approaches can provide an advantage over user-driven ones in work-order processing; however, specific data requirements need to be identified to collect and process the functional data needed while achieving more consistent and accurate results.
Design/methodology/approach
This paper presents the findings of an online survey conducted with facility management (FM) experts who are directly or indirectly involved in processing work orders in building maintenance.
Findings
The findings reflect the current practices of 71 survey participants on data requirements, criteria selection, rankings, with current shortcomings and challenges in prioritizing work orders. In addition, differences between criteria and their ranking within participants’ experience, facility types and facility sizes are investigated. The findings of the study provide a snapshot of the current practices in FM work order processing, which aids in developing a comprehensive framework to support data-driven decision-making and address the challenges with UDAs.
Originality/value
Although previous studies have explored the use of selected criteria for processing and prioritizing work orders, this paper investigated a comprehensive list of criteria used by various facilities for processing work orders. Furthermore, previous studies are focused on the processing and prioritization stage, whereas this paper explored the data collected following the completion of the maintenance tasks and the benefits it can provide for processing future work orders. In addition, previous studies have focused on one specific stage of work order processing, whereas this paper investigated the common data between different stages of work order processing for enhanced FM.
1. Introduction
The Facility Management-Vocabulary standard defines facility management (FM) as an:
[…] organizational function which integrates people, place, and process within the built environment with the purpose of improving the quality of life of people and the productivity of the core business (ISO 41011, 2017, p. 2).
FM integrates management practices with technical knowledge to plan, provide and manage effective built environment (Chanter and Swallow, 2008) with three main goals:
increasing productivity;
minimizing cost; and
providing information supporting strategic planning (Teicholz and Techolz, 2001).
FM practices constitute 60% of the building costs in its entire lifecycle (Guillen et al., 2016) with more than 65% of it being associated with maintenance costs (Chen et al., 2018; Becerik-Gerber et al, 2012). Operation and maintenance (O&M) practices generate a large amount of data requiring facility managers (FMs) to develop plans for collecting, processing and managing the data to support the organization’s goal. This is a key challenge in maintenance management (Chanter and Swallow, 2008), which may lead to poor facility performance when planned ineffectively (Besiktepe et al., 2020).
Prioritizing and processing work orders comprise a significant part of FM practices that generate a large number of orders in daily operations (Mo et al., 2017). User-driven approaches (UDAs) including manual work are currently more prevalent in processing work orders. Overcoming the challenges of managing work orders manually requires a team of specialized and experienced staff. Such approach brings additional challenges in FM practices that mainly include lack of consistency (Lukens et al., 2019) and subjectivity (Cao et al., 2015). To propose solutions addressing these challenges, more applied studies are required to reveal the industry practices, determine the gaps and challenges and identify best practices.
This paper provides a brief background on challenges and gaps in existing practice for processing maintenance work orders based on literature review and interviews conducted in prior research (Ensafi et al., 2023). It is important to acknowledge that the comprehensive literature review conducted in the previous study lays the background of the problem with revealing challenges in existing practices such as lack of data requirements and inconsistency in data collection, cognitive workload and biases and inconsistency among FMs. The paper presents the findings of an online survey questionnaire conducted with FM experts who are involved in processing work orders. The findings reflect the current practices of 71 participants on data requirements, criteria selection, rankings, with current shortcomings and challenges in prioritizing work orders. In addition, differences between criteria and their ranking within participants’ experience, facility types and facility sizes are investigated. Identifying and determining the criteria and their rankings in the FM practices can better support FM professionals in their strategic decision-making processes (Yoon et al., 2021; Besiktepe et al., 2020). Decision-making methods have been used to address some of the challenges with UDA to address the inconsistency and subjectivity. The paper concludes by describing the challenges associated with such methods and recommending the implementation of a more automated approach to address the challenges with (UDA). The findings of this study provide a snapshot of the current practices in FM work order processing, which aids developing a comprehensive framework to support data-driven decision-making (DDD).
2. Background
Every facility receives multiple work orders daily, with different severity and criticality levels. It is crucial to assess and prioritize the maintenance tasks as inefficient prioritization can lead to additional cost or equipment failure. Identifying the root causes and determining the resources required to address the work orders require multiple trips to the targeted area (McArthur et al., 2018). Studies have presented 50 billion dollars cost due to unplanned downtimes in manufacturing industry (IndustryWeek and Emerson, 2023). Furthermore, lack of proper prioritization can lead to work order backlog and deferred maintenance, which may result in significant costs (Gocodes, 2022). Priority management is the allocation of resources in a specific order to respond to operational pressure or customer service supporting the economic and strategic goals of an organization (Westbrook, 1994). Considering the complexity of these processes, standardization of data captured and definitions of maintenance tasks for creating, planning, scheduling and executing maintenance work orders are critical.
A current study by Ensafi and Thabet (2021) identified that most facilities have not developed specific data requirements for maintenance work orders, identified issues and information collected following maintenance task performed. As a common practice, processing work orders through phone calls and e-mails is prone to errors and inconsistency as the result of different levels of experience and knowledge, as well as judgment and biases of staff (Cao et al., 2015; Schwenk, 1985). Paper-based information is still used for transferring information, although maintenance management systems are available (Cheng et al., 2020). On the other hand, although some facilities are using maintenance management software, the data requirements have not been explored and determined in detail. Required data are not comprehensively collected and the collected data cannot be used effectively. In addition, there is a need to define data format requirements. Facilities collect information in textual and descriptive format which makes it more complicated to benefit from DDD.
Moreover, the criteria used by different individuals in the same facility are not consistent as there are no specified requirements for necessary criteria and their ranking for prioritizing maintenance work orders. Facilities staff use their knowledge, experience and familiarity with the facility to process and prioritize work orders. This common practice highlights the issue with the impact of experience, level of knowledge and biases (Schwenk, 1985). In addition, processing large amounts of maintenance work orders can lead to cognitive workload negatively impacting the performance of staff (Hollnagel and Woods, 2005). Studies have highlighted operator error and lack of time for maintenance as main reasons for unscheduled equipment downtimes (FinancesOnline, 2022). Delays in processing can lead to asset failure or downtimes negatively affecting the cost of O&M while reducing the safety and satisfaction of occupants (Hong et al., 2020). A study conducted by Chan et al. (2021) have marked unplanned equipment breakdowns as the top five challenges faced by facilities.
To address some of the challenges in existing practices, research efforts or practitioners have considered assigning a priority number (e.g. 1 for immediate and 3 for long-term concerns) to assets to determine their relative importance (Teicholz and Techolz, 2001). However, based on results from the prior study (Ensafi et al., 2023), the criticality number assigned to assets is not robust and may differ from daily practices when assigned outside of the maintenance situation. Another issue with existing systems is that they are not updated and adjusted over time to provide a practical solution.
Lack of data requirement created challenges with capturing data as well. To benefit from the data collected, it should be captured consistently and continuously. Many facilities have not defined requirements to determine what information should be collected following the completion of maintenance tasks. Such approach leads to discontinuous data collection hindering the efficient use of the data collected. Also, in cases where facility staff collect information, the information collected does not include details that would benefit future maintenance work orders.
All the factors mentioned above highlight the significance of proper maintenance work order management to improve the facilities’ performance. Maintenance work orders play an important role in FM as they can be used to keep the records and history of maintenance tasks (Lavy et al., 2019).
3. Methodology
To study the benefits of implementing DDD methods for processing maintenance work orders, data collection from a wide range of facilities and individuals was needed. Therefore, a questionnaire was conducted to collect information about criteria selection and ranking for prioritizing work orders and data collected following the maintenance tasks among different facility types and sizes.
Based on literature review and interviews conducted in prior studies, this research focused on collecting data from a wider range of facilities to include wider range of perspectives and practices and perform a more comprehensive analysis. The following steps were taken to conduct this study.
Step 1: A questionnaire was prepared and conducted to collect information about industry practices in terms of work order processing.
In addition to the question structure and flow, reliability and validity are critical in the survey design for accuracy and consistency. To address reliability, facility information such as type and size of the facility as well as background information of the participants were requested. With this information, the consistency of results from different individuals and facilities can be explored to show the representation of the sample group in the entire population. Furthermore, such approach allows the categorization of results into groups to explore the patterns and relationships between different factors. Literature review and a pilot study was conducted to address the validity. In the pilot study, the questionnaire was reviewed by two FM experts involved in processing work orders with more than 20 years of experience in different facility types to confirm the clarity of language and the development of questions in addressing the intended aspects.
Step 2: Two types of analysis were conducted. First, a descriptive analysis was conducted to present the results of criteria selection and ranking and compare the results among different facility types, facility sizes and years of experience. This analysis also determined the common data used in different stages of work order processing. Second, an inferential statistical analysis using normalization and multiple linear regression was conducted to investigate the possible relations between facility types, facility sizes and years of experience with criteria selection and ranking.
3.1 Data collection (Step 1)
Based on a prior study (Ensafi et al., 2023), a questionnaire was prepared to collect information about the criteria and their associated rankings used by individuals in different facility types for prioritizing work orders. The questionnaire also collected information about data requirements for processing work orders. Using questionnaires allows collecting information from a wider range of individuals and facilities reflecting and representing more realistic and robust results.
The questionnaire was structured into the following question types:
Multiple choice questions for participants’ background (e.g. years of experience) and facility information (e.g. facility type).
Ranking questions for selecting and ranking the 25 criteria used for work order processing.
Multiple answer questions for selecting the information collected following the maintenance task.
Open-ended questions for providing additional key information used for processing work orders.
To conduct the questionnaire, an application was submitted to the institutional review board at Virginia Tech. To recruit participants involved in processing work orders, the questionnaire was posted online on the Qualtrics website. In the first phase, individuals from the International Facility Management Association were contacted through professional connections of researchers by e-mails. The questionnaire was posted on an online platform (LinkedIn), in the second phase.
A total of 95 responses were collected between August and November 2021, from which 18 responses were excluded as the responders were not directly nor indirectly involved in processing work orders. In addition, six responses were excluded as they did not complete the questionnaire. Subsequently, 71 responses were analyzed for this study. Even though the nationwide population of FMs was reported around 100,000 by US Bureau of Labor Statistics (2021) Statistics in 2021, the number of FM professionals directly involved in work order processing in each facility is very limited. Acknowledging the fact that the authors are not making inferences about the entire FM population based on the sample size, the results of this study reflect the valuable feedback of participants in FM profession.
3.2 Analysis (Step 2)
First, a descriptive analysis was conducted to present the overall results of criteria selection and ranking. Descriptive analysis included calculation of percentages, averages and frequencies. The results were then categorized by facility type, facility size and years of experience. Descriptive analysis has been proven to be used to gain insights from a wide range of survey data for the following reasons. It simplifies the information into key results using central tendency, frequency and percentile. Such information allows determining trends and patterns and their frequency within data. The takeaway results provided are concise and actionable, which can significantly enhance communication with stakeholders especially when combined with the visual representations. Descriptive analysis can highlight the possibility of relationships, which can be further investigated using other statistical methods. Furthermore, this method allows comparison between different groups to determine similarities and differences. The simplified results allow understanding the behaviors and preferences as well (Mishra et al., 2019; Thompson, 2009). Also, an inferential statistical (Haden, 2019) analysis using multiple linear regression was conducted to investigate the relation between years of experience, facility type and size with criteria selection and ranking. While descriptive analysis describes the characteristics of the data, inferential statistics allow predictions to support DDD. Inferential statistical analysis allows comparison between groups and provides meaningful insight into patterns and relationships especially when it is complicated to collect data from the entire population providing a more efficient method to study the characteristics of the population (Dawson, 2012). The survey structure allowed the participants to select different numbers of criteria to better understand their practices requiring the rankings to be normalized to enable comparison of criteria ranking. Therefore, the rankings were first turned into scores by reversing the priority number. In other words, if a criterion was ranked as one, the score for that criterion would have been 23(total number of criteria). Second, the average score and standard deviation were calculated to be used for log normalization of the score. Log normalization was used instead of normalization as the data included 0 scores (unranked criterion), which would lead to computational issues. Using log normalization allowed to mitigate the effect of outliers and reduce the range of values and uneven spread of data. In the final step, multiple linear regression was applied to the normalized score to conduct the comparison. Multiple linear regression is used for studying the strength of relationships between two or multiple variables and therefore was used to investigate the relations between years of experience, facility type, facility size with criteria ranking. Finally, the data was analyzed to determine the common data between information used for processing work orders and information collected following the completion of the maintenance tasks using average and distribution (frequency) of the criteria selected. Determining the common data allows investigating similarities in different phases of maintenance work orders for developing comprehensive data requirements.
4. Analysis and results
This section first provides the results of the survey questionnaire analysis determining participants’ demographic information. The overall criteria selection and ranking are then presented. The results of the analysis also show the comparison of criteria selection and ranking based on facility type, size and years of experience. The top nine criteria were selected for further analysis considering the average number of criteria selected by the participants was nine. Using an inferential statistical analysis, the possible relations between years of experience, facility type and size with criteria selection and ranking are presented. The last section of results presents the common data between different stages of work order processing.
4.1 Participants’ demographic information
The following sections provide descriptive information about participants’ level of involvement, years of experience, facility type, facility size and facility location. Based on the questionnaire results, more than half of the respondents have been directly involved in processing work orders.
As a result of the literature review as well as interviews conducted (Ensafi et al., 2023), it is revealed that years of experience helps with better knowledge of the facility hence positively impacting the work order processing. Participants were asked to indicate their years of experience in FM to identify its impact on the selected criteria and their rankings. While 2.8% of participants had less than one year of experience, 32.4% of participants had 10–20 years of experience.
Participants were also asked to provide information about their facility type and facility size (number of buildings) to determine the impact of facility type and size on the criteria selection and ranking. Most respondents worked in commercial facilities as well as educational institutions covering 60% of the total responses (Figure 1). The lowest response rates were from health care and industrial facilities.
Out of the 71 responses, 30 facilities reported providing services to more than 50 buildings (Figure 2). In addition, 40 participants indicated that they manage all the buildings within their facilities, whereas 31 indicated managing a range of 6–30 buildings in their facilities.
Although the majority of responses (71.8%) were received from individuals within the USA, there were few responses from other countries including Australia (1), Brazil (1), Canada (2), Iran (2) and Qatar (2).
4.2 Criteria selection
Figure 3 presents the frequency of each criterion based on the responses. The results indicate that the frequently used criteria were “level of severity (hazard)”, “availability of staff” and “severity of failure”. The least selected criteria were “remaining lifecycle”, “energy usage” and “distance from the facility”. On average, nine criteria were selected for processing work orders, whereas the criteria selection ranged from 3 to 23.
The additional criteria considered by facilities included: “type of work order”, “impact of failure on business operations and facility function”, “impact on occupant's mission”, “potential for immediate or long-term damage to facility if not prioritized”, “business impact”, “corporate policy in terms of preventative maintenance schedules”, “importance of the requester’s role (e.g. vice president)” and “importance of requester’s location”.
The average number of selected criteria by individuals in each category of experience were compared in Figure 4. Individuals with zero or less than one year of experience (four participants) were disregarded due to the small sample size. It was observed that the average number of criteria selected by participants increased with their level of experience.
Considering the validity of the analysis, with the small sample size of health care, industrial, residential, hospitality and aviation facilities (Figure 1), the average number of criteria selected based on facility types were only compared for commercial, educational institutions, mixed-used and government facilities (Figure 5).
To further study the impact of facility type, the distribution of criteria selection was compared for two facility types of educational institutions and commercial facilities as they had the highest representation in the questionnaire results with the same number of participants. As Figure 6 presents, the frequency of the selected criterion differed for some of the criteria including availability of staff, level of risk and occupants’ satisfaction.
The average number of criteria selected among facilities with different sizes did not vary (Figure 7). Considering the finding, it can be assumed that when the number of buildings increases, not all buildings are managed by the same person/team, which requires further investigation.
4.3 Criteria ranking
While the descriptive analysis presented the attributes of the data, inferential statistics allow testing the hypothesis regarding the possible relationship between different factors. A statistical analysis was conducted to compare the ranking among different participants, and to do so, the rankings were converted to scores and were normalized.
Considering the top nine (based on average number of criteria selected) criteria ranking, level of severity was selected by the majority of participants with the highest ranking. However, categorizing the responses by years of experience, facility type and facility size shows that there are differences between the criteria selection and ranking among participants (Table 1).
Considering the differences between the sample sizes and their limitation in this study, the authors implemented multiple linear regression, to study the possibility of significant correlation between the three factors: years of experience, facility size and facility type with criteria ranking. Based on the analysis, facility type had significant impact on ranking of some of the criteria including maintenance duration (Prob > F: 0.0375) and budget (Prob > F: 0.0276). Years of experience also had impact on criteria ranking including end date (Prob > F: 0.0001), maintenance duration (Prob > |t|: 0.0452 for one to five years of experience), failure frequency (Prob > |t|: 0.0122 for 10–20 years of experience) and severity of failure (Prob > |t|: 0.0438 for zero years of experience but familiar). However, the facility size did not greatly impact the ranking of criteria.
4.4 Data collected after performing the maintenance tasks
According to interviews conducted in previous studies by the authors (Ensafi et al., 2023), most facilities collect information after the maintenance tasks are completed; however, not all facilities benefit from the data collected for future work order processing. In addition, due to lack of information requirements, FM departments are not necessarily collecting information that can be useful for processing work orders. Therefore, an analysis was conducted to determine what information is collected after the maintenance tasks are completed and if there is any overlap between information collected and information used for processing work orders.
Among the 71 responses, two respondents marked that they do not collect information after performing maintenance tasks. On average, five criteria were frequently selected by respondents as information collected following the maintenance task completion (Figure 8). Among the criteria provided, total hours spent on the tasks and issue description were the most selected (58 and 55, respectively), whereas percentage completion of the tasks and difficulty with repair were the least selected (14 and 9, respectively).
Other criteria indicated by the respondents include asset information, contracts and issue resolution, safety and security information, complaints from maintenance personnel, obstacles encountered, probability of repeated incidents, deficiencies noted and recommendations and opportunities for improvements, quality control and quality assurance notes, who performed the work, description of work performed, who closed the work order, whether the work was completed on time or was overdue, photos and final completion survey sent to requestor to ensure satisfaction level.
4.5 Common data between work order processing and information collected following the maintenance tasks
Table 2 presents the criteria selected by participants for processing work orders versus the information collected following the maintenance tasks performed. Understanding and identifying information requirements can help with developing data requirements to collect useful data consistently and continuously after the maintenance tasks are completed. For example, although 20 participants highlighted the importance of maintenance difficulty for processing maintenance work orders, only nine participants selected the difficulty of repair as information collected following the maintenance tasks.
Determining common data among different stages of work order processing allows avoiding the collection of data that is not important as well as considering important information that can be used to enhance work order processing. For example, maintenance cost was considered by 27 participants for processing work orders, whereas the cost of material and resources was captured by 53 of the participated in the study.
Further data analysis can be conducted to identify possible influential factors. For instance, if the facility captures duration of maintenance task and the name of the person who performed the tasks, the data can be used to assess the performance of staff.
While the numbers in Table 2 are good indicators of conceptual similarities between information used in different stages of work order processing, further analysis was performed. For instance, the selection of “Availability of maintenance staff” criterion is almost equal to collection of “Number of Crew” following the maintenance tasks. However, out of 47 participants who selected “Availability of maintenance staff” criterion, 16 of them did not select “Number of Crew” information collection following the maintenance tasks. In addition, 15 of the participants who indicated collecting “Number of Crew” information, did not select “Availability of maintenance staff” for work order processing. Same differences were identified for other information included in Table 2. Such differences highlight the importance of identifying information requirements to collect data that is needed, while also using the beneficial data that is already being collected.
5. Discussion
Considering the impact of facilities management on organizational effectiveness, FM decisions leverage business success (Alexander, 2013). DDD is pivotal to ensure organizational efficiency, especially in dynamic environments such as FM, which includes high uncertainty (Paganin et al., 2021). Together with this perspective, the findings of the study support the fundamental layout of using DDD in the FM body of knowledge that reinforces the applied and practical focus of FM. Moreover, the exploratory nature of the research methodology, delivering reflections from the FM industry, underscores the criticality of comprehending practical applications together with theoretical management frameworks.
Although previous studies have explored the use of selected criteria for processing and prioritizing work orders, this paper investigated a comprehensive list of criteria used by various facilities for processing work orders. Based on the questionnaire results, the number of criteria selected varied among the participants ranging from 3 to 23. Considering the range of number of criteria selected (3–23) and based on studies conducted on human cognitive limitations regarding cognitive workload and human limited cognitive capacity (Hollnagel and Woods, 2005), it would be complicated to consider more criteria at once. However, the results require further investigation to explore the reasoning behind the number of criteria selected while controlling other influential factors such as organizational goals.
This paper also investigated the impact of external factors such as facility type, size and years of experience on criteria selection and rankings. The average number of criteria selected among FM professionals varied based on years of experience and facility type. The results presented the selection of additional criteria by individuals with more experience. While increasing the number of criteria used for processing work orders can imply that more aspects are considered for prioritizing alternatives by individuals with high level of experiences, the results need to be validated by considering the impact of other factors such as organizational goals, strategic plans, maintenance budget, etc. The variety in number of criteria selected across individuals highlights the need for identifying the optimum number of criteria to collect comprehensive information while avoiding excessive information collection. Furthermore, there should be a balance between the processing time and the quality of results. For instance, it is important to explore if using more criteria will lead to longer processing time and what would be the impact of longer processing times, their costs, labor hours and planning efforts to determine the benefits and challenges.
Considering the overall criteria ranking, level of severity, severity of failure and availability of maintenance staff, respectively, were ranked as the top three criteria selected by the participants. The rankings were compared among different facility types, sizes and years of experience and the results presented differences between the categories. Such differences could be an indication of differences in organizational values among different facility types leading to different criteria selection and ranking, which require further investigation. It is important to determine such factors as a single list of criteria, which would be considered as a generic list; however, it is critical to have criteria for specific building types and needs. The list of criteria for prioritizing work orders should have flexibility to be applicable to various facilities considering the differences in their functionality, size and goals.
The survey structure let the participants select and rank their preferred number of criteria; in other words, participants did not have to select and rank all criteria provided. Such option impacted the statistical analysis. Although the results of the regression analysis marked the possible correlation between external factors and criteria selection and ranking, these results were impacted by the sample sizes in each category. Therefore, the authors have focused on presenting the results with descriptive techniques. To further study the impact of each factor and validate the results of this research, bigger sample sizes are required. In addition, larger sample sizes will allow investigating the impact of multiple factors (e.g. facility type and size) simultaneously.
Most of the previous studies are focused on the processing and prioritization stage while this paper explored the data collected following the completion of the maintenance tasks in existing practices and the benefits it can provide for processing future work orders. In addition, previous studies have focused on one specific stage of work order processing, whereas this paper investigated the common data between different stages of work order processing considering the connection between different stages for enhanced FM. Although most facilities collect information after the maintenance tasks are completed, not all of them benefited from the data collected for future work order processing. In other words, while participants highlighted the importance of some of the criteria for processing work orders, they did not collect that information following the maintenance tasks completion. On the other hand, the data they collected following the maintenance tasks were not necessarily beneficial or used for processing maintenance work orders. Understanding common information between different stages of processing work orders can help with developing comprehensive data requirements to address the existing challenges with strategic decision-making and can enforce what data should be collected and in what format. FM professionals should consider the big picture and the connection between different stages of work order processing. Such approach helps with collecting sufficient information and eliminates data duplication to save time, enhance the quality of services and support the future decision-making processes. Insufficient information can force the operator to interpret based on limited amount of information while providing excessive information may also lead to coping strategies.
Studies conducted on prioritization have used decision-making methods such as priority criterion (Yusof et al., 2012), technique for order of preference by similarity to ideal solution (Shyjith et al., 2008), analytic network procedure (Chemweno et al., 2015) and analytical hierarchy process (Ohta et al., 2018) to assign weights to criteria for prioritizing a set of alternatives. Although decision-making methods can be used to enhance the consistency in prioritizing a set of alternatives, there are some challenges and gaps in their application. First, the alternatives usually have multiple attributes (e.g. dividing cost into purchased cost, labor cost, maintenance cost), consequently, considering one attribute for each alternative is not realistic. On the other hand, due to limited human cognition capacity and as a result of information overload, humans cannot reliably compare multiple pairwise alternatives with different attributes simultaneously. Therefore, their decision-making is negatively impacted requiring systematic decision-making for complex decisions with conflicting criteria (Dixit, 2018; Hollnagel and Woods, 2005). Second, the comparison of alternatives is different and more realistic when it is performed during the actual maintenance work order prioritization compared to when it is performed outside of maintenance context (Ensafi et al., 2023). For instance, distance was marked as one of the least important criteria for work order processing. While the level of importance for distance is correct, compared to other criteria such as risk level, previous study by the authors have revealed that individuals consider distance with higher priority in some cases where it increases the convenience of the staff (e.g. assigning work orders located in a shorter distance to crews). Third, humans do not assess the pairwise comparison matrices in a consistent manner as they use a combination of quantitative and qualitative analysis (Dixit, 2018). Fourth, possible influential factors such as years of experience should be identified and considered when prioritizing a set of alternatives to increase consistency and remove the impact of judgments and possible biases. Fifth, the weights estimated for different attributes are not updated over time unless the rankings are investigated repeatedly (Ensafi et al., 2023). Sixth, some staff may still rely on their knowledge and judgement instead of following the prioritization strategy.
As large amount of data is generated by various assets within the facilities and are available to FMs, the concept of big data can assist the facilities with extracting meaningful information and patterns. Such approach addresses some of the challenges with user-driven decision-making but requires development of data requirements as well as continuous and consistent data collection.
6. Conclusion
This study bridges the gap between theory and practice by comparing practices among different facilities supporting the development and optimization of FM solutions. This paper presented the results of a questionnaire conducted to investigate the overall criteria selection and ranking and to compare the results among different facility types, sizes and years of experience. The questionnaire allowed validating the criteria identified in prior study while covering a wider range of practices and perspectives. The differences in criteria selection and ranking among the three categories highlight the impact of factors that should be considered when developing data requirements to enhance consistency in work order processing.
As presented in results, the average number of criteria selected by participants was nine, whereas this number varied among participants. This highlights the need to explore the efficient number of criteria to be used based on human limited cognitive capacity and effective pairwise comparison. Such approach also enhances consistency among FMs with different years of experience.
The results of this research provided details on data collected following the maintenance tasks and compared the data with the data used for processing work orders for each facility. Such approach allowed gaining an in-depth insight into the data overlap between different stages and identifying the useful information collected by the facilities. The results can also be an indication of lack of awareness regarding the data stored in the system that can be used by FMs for processing work orders. Disconnection between different stages and the data collected force facilities to use multiple platforms for different phases of processing, creating challenges with data management and interoperability. Increasing the understanding of FMs in terms of the connection between different stages of work order processing and the data collected and stored within each stage can help with their strategy planning and data requirement development.
A descriptive and an inferential statistical analysis were conducted to gain insight from the data collected and investigate the possible impact of different factors on criteria selection and ranking. The results presented possible relation and impact of years of experience and facility type on data collection and ranking. Considering impactful factors such as facility type will allow developing practical requirements that can be adapted based on organizational characteristics. The paper discussed the shortcomings of implementation of decision-making methods, which are used to address some of the challenges with UDAs such as inconsistency. The paper concluded with suggestions on using data-driven methods.
The results of this study have multiple implications. Although technology has provided many opportunities for optimization and improvements, FMs have faced obstacles such as lack of data quality to adopt them. The results of this study can help with development of the data requirements to collect high-quality data for informed decision-making. Such approach reduces the cost of O&M and improves productivity and occupants’ satisfaction. The result of the descriptive analysis is straightforward, contains the key outcomes and is actionable enhancing the communication with stakeholders to improve existing processes. The results have also shown the connection and impact of different stages of work order processing on one another which can further support the development of data requirements and strategic plannings to improve future work order processing. Furthermore, there is a need for developing flexible solutions that can be adjusted based on facility features such as type and size to provide long-term practical solutions. In conclusion, this study in addition to the previous study conducted by the authors provides fundamental structure for developing and optimizing FM systems to better support work order processing and DDD.
The limitations of this study were small sample sizes in some categories such as one year of experience or residential facilities, inequivalent sample sizes and differences in number of criteria selected among participants. Although the data collected from the questionnaire was sufficient to perform a descriptive analysis and present the differences among different categories, more responses are needed to have larger sample sizes (e.g. health-care facilities) to validate the statistical analysis and the possible correlation between the factors discussed. The results of this study provided a preliminary understanding with a snapshot of how work order processing works from an industry perspective. Future studies should focus on collecting more data points for making a general inference.
Due to the challenges with decision-making methods and to benefit from historical data collected, the authors are planning to implement DDD, specifically neural network, to automate the prioritization of future work orders. Machine learning (ML) methods and algorithms have been used by different researchers in various fields including construction (Ensafi et al., 2022) and FM (Zarindast and Wood, 2021; Canizo et al., 2017) to address the challenges with existing practices. Having access to historical maintenance data and work order history can assist FMs with processing future work orders as well as determining correlations between different factors. The use of ML allows covering more criteria without resulting in cognitive workload and coping strategies. Using neural networks, the optimum number of criteria can be determined, and the importance of each criterion can be estimated based on previous knowledge and approaches used for processing work orders. In addition, using methods such as ML allows automatic updates based on new inputs, which leads to more practical solutions over time.
The authors would like to sincerely thank John Mackay, Mani Ardalan Farhadi and Diane Coles Levine for supporting this research and helping in distributing the questionnaire.
Number of responses by facility type
Number of responses based on facility size
Frequency of criteria selection
Average number of criteria based on years of experience
Average number of criteria based on facility type
Selection percentage based on facility type
Average number of criteria based on facility size
Data collected following the maintenance tasks
Top ranked criteria based on years of experience, facility type and facility size
| Criteria | Overall ranking | Years of |
Facility type | Facility size |
|||
|---|---|---|---|---|---|---|---|
| 1–5 | Over 20 | Educational institution | Commercial | 1–5 | 11–50 | ||
| Type of building | 6 | 7 | 7 | 4 | 6 | 7 | 6 |
| Type of space | 9 | 9 | – | 5 | – | – | – |
| Availability of maintenance staff | 3 | 4 | 4 | 2 | 5 | 8 | 7 |
| Budget | – | – | 8 | – | – | – | – |
| Maintenance cost | – | – | – | – | – | – | 9 |
| Level of severity (hazards) | 1 | 1 | 1 | 1 | 1 | 6 | 1 |
| Failure frequency | – | – | – | 9 | – | – | – |
| Severity of failure | 2 | 3 | 2 | 3 | 4 | 2 | 2 |
| Level of risk | 5 | 5 | 5 | 8 | 3 | 4 | 3 |
| Availability of resources | 7 | 6 | 7 | 7 | 3 | 8 | |
| Indoor environmental quality | – | – | 9 | – | – | 9 | – |
| Association between different equipment/system | 8 | 8 | – | – | 9 | – | 5 |
| Occupants’ preferences\satisfaction | 4 | 2 | 3 | 6 | 2 | 1 | 4 |
| Codes and regulations | – | – | 6 | – | 8 | 5 | – |
Common data
| Selection |
Criteria for processing work orders | Information collected following maintenance task | Selection |
|---|---|---|---|
| 47 | Availability maintenance staff | Number of crew members | 46 |
| 37 | Availability of resources | Quantity of spare parts required | 29 |
| Material used | 1 | ||
| 20 | Indoor environmental quality | Safety and security information | 1 |
| 49 | Level of Severity (hazard) | Issue description | 55 |
| 27 | Maintenance cost | Material/resources cost | 53 |
| 25 | Maintenance duration | Duration of task (start and end date) | 53 |
| Total hours spent on the task | 58 | ||
| Number of visits to complete the task | 20 | ||
| Percentage completion in each visit | 14 | ||
| 25 | Failure frequency | Likelihood of repeat incidents | 1 |
| 20 | Maintenance difficulty | Difficulty with repair | 9 |
| Obstacles encountered | 1 | ||
| Complaints from maintenance personnel | 1 | ||
| 18 | End date | Whether the work was completed on time or was overdue | 1 |
| 40 | Occupants' satisfaction/preference | Client/occupants' satisfaction | 4 |
| 31 | Codes and regulation | Contracts and issue resolution | 3 |
| 32 | Systems association | Any tracked assets the work was associated with | 1 |
| 25 | Maintenance type, type of work | What was done to complete (work description) | 2 |
| 28 | Details of problem, exact location of issue or asset | Issue description | 55 |
Source: Created by authors
© Emerald Publishing Limited.