NCM COMPANY’S SYSTEMS, EMPLOYEE AWARENESS AND READINESS FOR INTEGRATING AI-DRIVEN PREDICTIVE MAINTENANCE IN COMPOSITE PANEL PROCESS: BASIS FOR STRATEGIC IMPLEMENTATION

ABSTRACT

This thesis investigates how NCM Company handles its maintenance responsibilities in composite panel process, as well as the perceived levels of awareness and readiness of employees for integrating Artificial Inteligence to predictive maintenance system of the company. Specifically focused on the maintenance activities such as spotting anomalies, predicting equipment failures, estimating how long components will last, and keeping track of production efficiency.  The researcher collected data from 53 selected managers, supervisors, engineers, and technicians. The survey findings provided valuable insights into the existing levels of awareness and readiness crucial for the successful integration of AI into the company's predictive maintenance system. The research stressed the importance of fostering awareness of AI predictive maintenance for effective integration into the workforce. The results highlighted the need for specific training, especially for lower-level employees, considering the different level of awareness within the organization. The study emphasized the significance of organizational initiatives, such as orientation programs in enhancing awareness and cultivating a technologically proficient workforce. In conclusion, the research underscored the proactive necessity of closing the awareness gap and providing employees with comprehensive knowledge of AI-driven maintenance. Recommendations included further assessing AI integration, understanding employee reactions, and employing diverse research methods for a holistic understanding.

INTRODUCTION

Nature and Scope of the Problem Investigated   

Maintenance practices are crucial for achieving high Overall Equipment Effectiveness (OEE), but traditional methods, such as scheduled or reactive maintenance, often lead to unforeseen breakdowns, delays, and increased downtime. This approach can be costly and inefficient, with maintenance activities accounting for up to 40% of total production costs and a significant portion spent unnecessarily due to various failures. The lifespan and availability of machine parts, particularly on laminating press equipment crucial for composite panel production, pose significant challenges, leading to extended downtime and reduced Overall Equipment Effectivity. The lack of spare parts availability affects both machine performance and maintenance lead time, further aggravating downtime and compromising equipment effectiveness. NCM Company's reliance on manual data input for maintenance tracking exacerbates these challenges, hindering proactive maintenance measures and decision-making. Recognizing the importance of innovation and digital transformation, The researcher explores the integration of artificial intelligence (AI) into its predictive maintenance system as a solution to improve equipment reliability and reduce downtime. AI's ability to automatically detect machine failures and predict part wear offers timely insights, while transitioning to automated data collection streamlines maintenance operations and enhances predictive capabilities.

Research Problems and Objectives  

This research determined and provided answers to the following research problems and objectives:

• The existing NCM company’s predictive maintenance system for Machine on company’s composite panel processc.
• The level of agreement on awareness employees in NCM company regarding the application of Artificial Intelligence in predictive maintenance system in their composite panel process.
• The level of agreement on readiness of the employees in NCM company regarding the application of Artificial Intelligence in predictive maintenance system in their composite panel process.
• The plans and strategies that can be recommended to embrace AI-driven enhancements in predictive maintenance system of the company.

Research Framework   

The research's conceptual framework directed the investigation of vital aspects in NCM Company's transformation, particularly in AI-driven predictive maintenance for the Composite Panel Process. It included evaluating systems, employee awareness, readiness, and AI adoption strategies. The framework outlined a systematic progression from system assessment to AI implementation, assessing employee awareness levels and readiness for anomaly prediction, failure estimation, and production efficiency metrics. Emphasizing alignment with operational goals, fostering a supportive workplace culture, and implementing strategic plans were crucial for effective AI integration and operational efficiency improvement.

Figure 1. Conceptual Framework of the Study

Research Significance 

This research contributed to the use of Journey of Transformation Model (Anderson, et al., 2010) in adapting new technologies, such as AI in predictive maintenance system in organizations. This Journey of Transformation Model served as a guiding framework for this research by emphasizing the importance of assessing the current state, articulating a vision, engaging employees, implementing changes, and ensuring ongoing feedback and improvement. The application of this model helped structure the research findings and recommendations in a way that supported a successful and sustainable transformation of predictive maintenance practices in NCM Company and similar industries.

Philosophical Lens   

This research adopted a positivist philosophy, emphasizing empirical observation and scientific methods to explore AI-driven predictive maintenance. August Comte's positivism guided the objective gathering of empirical data, analysis, and drawing conclusions. Positivism is a philosophical stance emphasizing the objective and empirical observation of phenomena, seeking to establish scientific principles through systematic and rigorous methodologies. This way of thinking is in line with a quantitative research approach, which highlights the importance of measuring and analyzing things we can observe to reach objective conclusions. The study-maintained impartiality and credibility by employing non-intrusive methods outlined in research ethics principles. Top of Form   

Scope and Limitations   

This study was limited to the composite panel fabrication process of NCM Company, located in the First Industrial Park in Batangas City, Philippines. to the composite panel fabrication process of NCM Company situated in the First Industrial Park in Batangas City, Philippines. The research gathered data from operational units and other pertinent departments intricately involved in the company's panel fabrication process. The scope of the research is confined to exploring the level of understanding and readiness among employees, overlooking crucial aspects such as the preparedness of data systems and the infrastructure supporting them. This limitation implies that the broader organizational ecosystem, encompassing data management and technological infrastructure, remains unexplored within the confines of this study. Furthermore, this research was limited to 53 employee respondents approved by the HR of the chosen aerospace manufacturing company. The employees involved in this research are Managers, Supervisors, Engineers, and Technicians directly involved in composite panel fabrication processes within NCM company.

Definition of Terms   

Anomaly and Failure Prediction. This involves the use of data analysis and models to predict irregularities or malfunctions in systems or machinery before they occur. 

Artificial Intelligence Concept. This involves selecting the most suitable artificial intelligence concept for potential enhancements that can be integrated into a process. 

Composite Panel.  Refers to sandwich structures composed of multiple layers bonded together to create multi-layer sheets, cored laminates, or industrial structural panels. 

Overall Equipment Efficiency (OEE). A metric used to measure the effectiveness and performance of manufacturing processes or individual pieces of equipment.   

Predictive Maintenance (PdM). A technique that uses data analysis tools to detect anomalies and potential defects in equipment and processes before they result in failure. 

Production Efficiency Metrics. Refer to measurements and indicators used to evaluate the effectiveness and performance of manufacturing processes or systems. 

Remaining Useful Life.  Refers to the estimated operational lifespan of a system or asset before it becomes unreliable or inefficient.

Review of Pertinent Literatures   

In the field of engineering, systems and structures degrade unpredictably due to time and external factors like sudden shocks (Caballe & Castro, 2017), leading to failures ranging from minor jams to critical breakdowns with safety risks and shutdown potential. Aging worsens issues, prompting significant maintenance investment. Large organizations typically employ engineering experts for preventive maintenance, with practices varying, including predictive maintenance integration to enhance decision-making. Emphasis is placed on overall equipment effectiveness (OEE) beyond availability, considering performance and quality (Forester et al., 2019). Organizations are increasingly adopting predictive maintenance (PdM) strategies to tackle maintenance challenges, monitoring equipment health in real-time and using data analytics to predict and prevent failures (Moini, 2020). PdM enables proactive maintenance, reducing downtime, and optimizing equipment performance by identifying potential failure patterns. The integration of artificial intelligence (AI) enhances PdM capabilities by analyzing large datasets and identifying subtle patterns overlooked by human operators (Nguyen et al., 2020). AI-driven systems provide accurate predictions, actionable insights, and optimized maintenance schedules, reducing costs and maximizing equipment reliability. However, successful AI adoption requires assessing organizational readiness to embrace AI-related technologies (Alsheibani et al., 2018). Conducting a thorough AI readiness assessment helps identify potential gaps and positions organizations for effective AI implementation, enhancing preparedness, and optimizing resource utilization. The successful adoption of AI depends on various factors, including the acquisition of enhanced technical skills and robust top management support (Achmat and Brown, 2019; Alsheibani et al., 2020; Jöhnk et al., 2021; Pumplun et al., 2019). Historical perspectives, such as Gill's (1995) observations on the loss of developers’ post-adoption in early expert systems, highlight the enduring importance of skill development and managerial backing in AI integration. Recognizing the evolving impact of AI on individuals' motivation and proficiency underscores the significance of skill enhancement and leadership support in navigating AI adoption challenges. The operational dynamics of socio-technical systems are often conceptualized through frameworks like Leavitt's (1964) "diamond" model, which encapsulates the interplay among technology, people, tasks, and structure. This model emphasizes the intricate relationships within organizational systems and provides a foundational understanding of their multifaceted nature. In contemporary AI adoption projects, the "golden triangle" model, comprising People, Processes, and Technology, emerges as a fitting framework for analyzing socio-technical dynamics. This model acknowledges the interconnectedness of human elements, operational processes, and technological infrastructure, aligning with the multifaceted considerations prevalent in AI adoption literature. Research supports the relevance of each element within the "golden triangle" model, emphasizing AI's impact on individuals, the importance of well-defined processes, and the alignment of technology with business needs (Frey and Osborne, 2017; Klumpp, 2018; Makarius et al., 2020; Saari, Kuusisto, & Pritikanga’s, 2019; Coombs, Hislop, Taneva, & Barnard, 2020; Kolbjørnsrud et al., 2016; Tarafdar et al., 2019). This integrative model offers a comprehensive lens for examining the multifaceted dynamics of socio-technical systems, providing insights into the interconnected factors shaping successful AI adoption.

METHODOLOGY

Research Design 

This research used a quantitative descriptive design which supported the defined objectives of this research. The primary data were composed of managers, supervisors, engineers, and technicians’ awareness and readiness regarding adoption of AI in predictive maintenance system. 

Research Locale 

The research was conducted in a Composite Panel Process at NCM Manufacturing  Company   situated   in   Region  IVA ( CALABARZON ),  Batangas, Philippines. The participants held different position from management to rank and file employee’s positions from different departments, including program management, manufacturing, engineering, maintenance, and quality which has direct involvement on composite panel process of the NCM company.

Population and Sampling Design   

The participants were employees of legal working age currently working at the NCM manufacturing company in the Batangas, Philippines. They represented diverse departments including maintenance, operations, engineering, quality, and program management, occupying positions such as technicians, engineers, supervisors, and managers. In total, 53 employees involved in panel fabrication, responsible for both production and maintenance tasks, formed the study's population.   

Research Instruments   

Quantitative research design was used for this research. A standardized instrument, via a survey questionnaire, was used to collect the primary data. The survey was structured into 3 distinct parts with the primary  aim  of  assessing   the awareness and  readiness levels of the employees regarding the implementation of AI in the company's predictive maintenance system.   

Data Gathering Procedure   

The primary data used on this research were collected via a survey questionnaire. The necessary permissions for the study involving company employees were obtained by the researcher. The data gathering procedure for this research covered the steps from requesting approval from the Human Resources department to collecting data from the respondents. The request for approval was communicated to both the Human Resources and Legal departments. The collected data was analyzed using descriptive statistical techniques, such as percentage and frequency distribution.

Management and Treatment of Data   

To analyze the data obtained from the survey, descriptive statistics were employed. This statistical method helped produce a comprehensive summarization of the survey responses, aiding in the identification of awareness and readiness levels among the employees of NCM manufacturing company.

RESULTS AND DISCUSSION  

Existing NCM company’s predictive maintenance system for Machine on composite panel fabrication process. 

NCM Company's predictive maintenance systems heavily relied on manual methods, with maintenance teams gathering data, analyzing trends, and making predictions based on their experience. This approach demanded proactive maintenance, involving direct observations, routine inspections, and periodic measurements to monitor industrial equipment health. Technicians played a crucial role in this process, ensuring firsthand understanding of asset conditions and detecting subtle signs of deterioration. However, manual data gathering posed challenges in terms of time, accuracy, and continuous monitoring capability. Table 1 provides a glimpse into NCM Company's condition-based monitoring process, underscoring the vital role technicians play in hands-on data collection. Technicians actively gather data on various aspects, including conditions, performance metrics, and deviations from specified parameters. These data are instrumental in identifying potential issues, conducting trend analyses, and informing maintenance decisions.

Table 1. Data Collection: Predictive Maintenance Checksheet 

The highlighted values in the table serve as more than just representations; they serve as crucial indicators signaling an immediate need for maintenance activities such as cleaning or parts replacement. These indicators highlight points where the well-being of components or their specifications has fallen below recommended values, necessitating urgent attention to ensure optimal performance and prevent potential issues from arising proactively. This symbiotic relationship between manual data collection and critical value indicators underscores the comprehensive methodology employed in maintenance practices within NCM's composite panel fabrication process. The company also employs IoT communication protocols for overseeing Production Efficiency Metrics, utilizing a lightweight protocol to gather data from machines, including Programmable Logic Controllers (PLC) in the panel manufacturing process. Despite this advanced technology, the primary role remains as a monitoring system for production KPIs. The operational framework relies on manual intervention by technicians for problem description and parameter input. While IoT streamlines data collection, decision-making and problem identification heavily depend on human input.

Level of awareness of the employees in NCM company regarding the application of Artificial Intelligence in predictive maintenance system  

Based on the collected data shown in table 2, on average, 64% of all employees had a high level of awareness, 13% had a moderate level, and 24% had a low level of awareness regarding AI in predictive maintenance systems. Managers exhibited a high level of familiarity with AI-driven predictive maintenance, indicating a strong understanding among this group. Managers demonstrated a high level of familiarity with AI-driven predictive maintenance, indicating a strong understanding within this group. However, supervisors displayed a mixture of high and moderate to low awareness levels, indicating a knowledge gap compared to managers. This highlights the necessity for targeted training to prepare supervisors for effectively guiding teams in AI-related tasks, especially if integration into the company’s predictive maintenance system is planned. Engineers showed a significant percentage with moderate and low awareness. Technicians, on the other hand, presented a considerable challenge, as most fell into the moderate and low awareness categories. Overall, the data from Table 2 underscores variations in awareness levels across different employee roles. While managers generally exhibit higher awareness in some aspect, supervisors, engineers, and technicians show significant percentages with moderate to low awareness, indicating areas for improvement.   

Table 2: Level of awareness of the employees regarding the application of Artificial Intelligence in predictive maintenance system

Level of readiness of the employees in NCM company regarding the application of Artificial Intelligence in predictive maintenance system. 

The readiness of employees at NCM Company to adopt Artificial Intelligence (AI) in predictive maintenance for the composite panel process indicates their preparedness to leverage AI technologies effectively.

Table 3 provides an overview of this readiness, revealing enthusiasm among participants to integrate AI into current manufacturing practices. On average, 92% of employees demonstrate a high readiness level, 8% show moderate readiness, and none exhibit low readiness. These findings reflect a positive organizational culture and a proactive approach toward embracing technological advancements to enhance maintenance efficiency and effectiveness.

Overall, the data indicates a high level of readiness among employees for adopting AI-driven predictive maintenance systems.With high readiness levels across managerial, supervisory, engineering, and technical roles, the organization is positioned to effectively integrate AI technologies into maintenance processes potentially leading to enhanced operational efficiency and performance.

Table 3: Level of readiness of employee regarding the application of AI in   predictive maintenance system

Identification of plans and strategies that can be implemented to embrace AI-driven enhancements in predictive maintenance.   

The human resource management function of an organization has an important role to play in effectively incorporating AI at work (Lawler & Elliot, 1996; Strohmeier & Piazza, 2015). Integrating the processes of human resource management along with artificial intelligence can generate additional benefits for an organization (Minbaeva, 2020), such as improved managerial decisions (Liboni, Cezarino, Jabbour, Oliveira, & Stefanelli, 2019), faster and more effective employee recruitment processes (Reilly, 2018), better learning at work (Hamilton & Sodeman, 2020), employee engagement (Tripathi, Ranjan, & Pandeya, 2012), and employee retention (Samarasinghe & Medis, 2020). Managers lead the integration of AI-driven enhancements in predictive maintenance by initiating educational frameworks, consulting external AI experts, and overseeing tailored educational programs. They promote continuous learning and innovation, championing AI adoption and allocating resources for training. Supervisors ensure practical exposure to AI tools through hands-on pilot projects and department-specific training sessions. They foster a culture of transparency and open communication to address challenges. Engineers and technicians undergo customized training, participate in pilot projects, and collaborate to identify AI integration opportunities, fostering adaptability and contributing insights for refining AI applications aligned with maintenance needs.

RESEARCH IMPLICATIONS

Summary of Findings   

NCM has an established predictive maintenance system, However, the system currently operates without fully automation, predominantly relying on manual data collection. This reliance on manual processes may be contributing to sub optimal production key performance indicators (KPIs), particularly affecting the Overall Equipment Effectiveness (OEE). The survey findings reveal differing awareness levels among employees regarding the application of AI-driven Predictive Maintenance. While most grasp AI concepts, specific awareness about AI-driven predictive maintenance within the system remains moderate, with a notable percentage showing low awareness, especially among rank-and-file staff. On the other hand, the surveys indicate a remarkable readiness among employees to adopt AI technologies. 

Derivable Conclusions from Research Data 

The existing maintenance system at NCM Company, although adept at predicting anomalies and estimating component life, currently relies on manual processes, potentially impacting Overall Equipment Effectiveness (OEE). In the rapidly evolving technological landscape, where automation and artificial intelligence (AI) played crucial roles in industrial processes, manual approaches may have fallen behind. This suggested that NCM Company could have faced difficulties in keeping up with industry advancements, potentially affecting its capacity to enhance equipment performance and reduce downtime effectively. The paradox of low awareness and high readiness among employees highlights the need for strategic action from organizations. While it's vital to address awareness gaps to ensure a thorough understanding of AI applications, organizations can benefit from employees' existing readiness. Implementing targeted programs such as awareness campaigns, training sessions, and educational initiatives can capitalize on this readiness. By investing resources into these efforts, organizations can equip their workforce with the necessary knowledge and skills to effectively utilize AI in predictive maintenance, thus maximizing its potential benefits. It's important to recognize that while high readiness levels among employees provide a valuable foundation for AI adoption, they are not sufficient on their own. The presence of awareness gaps indicates that readiness alone cannot guarantee seamless integration and effective utilization of AI in maintenance practices. 

Research and Policy Recommendations   

To enhance awareness of AI integration in predictive maintenance for composite panel systems, comprehensive orientation programs are vital, tailored to different employee levels. These programs should provide foundational AI knowledge, emphasizing its role in predictive maintenance, with interactive elements for engagement. Strategic awareness campaigns using multiple communication channels should emphasize the transformative potential of AI in maintenance processes. Visual aids, infographics, and success stories can make complex concepts accessible to all employees, addressing potential challenges, and showcasing success stories from other industries will ease the transition process. Securing top leadership commitment to champion the AI integration process is crucial. Involving leadership in communication efforts, emphasizing strategic alignment with organizational goals, and leveraging leadership influence will instill confidence and enthusiasm among employees. Quantitative research methods may have limitations; future studies should consider alternative methods, including qualitative or mixed methods research, to supplement the findings. Assessing employee awareness, perception, and readiness for AI may be susceptible to subjective opinions and biases. Complementing self-reported data with objective measures or alternative methodologies is recommended for enhanced reliability and accuracy. Finally, to advance understanding, future research should focus on comprehensive AI integration strategies, technological infrastructure assessment, and cross-industry comparative studies to identify best practices and obstacles.

WORK CITED

Achmat, L., & Brown, I., (2019). Artificial intelligence affordances for business innovation: A systematic review of literature. https://scholar.google.com/scholar.

Alsheibani, S., Cheung, y., Messom, C., (2018) Artificial Intelligence Adoption: AI-readiness at Firm-Level.

Cardoso, D., Ferreira, L., (2020). (PDF) Application of Predictive Maintenance Concepts Using Artificial Intelligence Tools. (researchgate.net).

Castro, I., Caballe, N., (2016). A condition-based maintenance of a dependent degradation-threshold-shock model in a system with multiple degradation processes. https://doi.org/10.1016/j.ress.2014.09.024.

Cheng, B., Lin, H., Kong, Y., (2023). Challenge or hindrance? How and when organizational artificial intelligence adoption influences employee job crafting – ScienceDirect.

Coombsa, C., Hislopb, D., Tanavec, S.K., Barnarda, S., (2020). The strategic impacts of Intelligent Automation for knowledge and service work: An interdisciplinaryreview.

Damant, L., Forsyth, A., Farcas, R., VOIGTLÄNDER, M., SINGH, S.; Fana, I.S.; SHEHAB, E.; (2021). Exploring the Transition from Preventive Maintenance to Predictive Maintenance Within ERP Systems by Utilising Digital Twins. DOI:10.3233/ATDE210095.

Frey, C.B., Osborne, M., (2017). The future of employment: How susceptible are jobstocomputerization? https://www.sciencedirect.com/science/article/abs/pii/S0040162516302244.

Saari, L., Kuusisto, O., & Pirttikangas, S. (2019). AI Maturity Web Tool Helps Organizations Proceed with AI. VTT Technical Research Centre of Finland. VTT White Paper https://doi.org/10.32040/Whitepaper.2019.AIMaturity.

Strohmeier, S., Piazza, F., (2015). Artificial Intelligence Techniques in Human Resource Management—A Conceptual Exploration.