Skill set issues in aircraft maintenance from industrial revolution 4.0 context: A document analytics survey

2.1 Level of aviation maintenance tasks

Aircraft maintenance technicians work in a number of highly specialized fields, such as aircraft construction and engine maintenance, as well as system maintenance, such as instruments, navigation and communication components. These technical crews are in charge of making sure that aircraft fly safely and efficiently. These experts maintain, service, repair, and maintain aircraft parts and systems in order to ensure the aircraft’s complete reliability and safety. Aircraft maintenance is also a fast-paced industry. In the future, aircraft maintenance will continue to evolve [6]. This is due to the introduction of new aircraft designs and materials, as well as the interaction of complex advanced technologies such as mission control computers, fly-by-wire systems with scheduled maintenance systems such as hydraulics, flight controls, and propellers or blades systems, microsensor networks interconnected, intelligent and intuitive presentation of data in remote operations [65].

In general, aircraft maintenance technicians perform a variety of maintenance tasks, which are supported by three-tiered maintenance plans as follows:

1st level maintenance (Organizational level): This on-site maintenance check includes inspection, repair, and part replacement.

All aviation system maintenance can be classified as corrective, preventative, or surveillance, as well as scheduled, unscheduled, or condition maintenance.

2.2 Aviation maintenance and skillsets issues and incidents

Aircraft maintenance refers to the repair and service activities performed to keep the aircraft operational and reliable. Maintenance activities encompassing component or aircraft engine repair, inspection and problem-solving [7–9]. Aircraft maintenance is a critical task in ensuring flight safety throughout the aircraft’s life cycle [10–12]. Despite the fact that aircraft maintenance is considered a high-risk job task in aviation due to its critical impact on aviation safety, it still plays a significant role in aircraft accidents and incidents [13–16]. A couple of studies [17, 18] emphasised the importance of developing a culture of recognising, reporting, and learning from past maintenance errors in order to improve maintenance quality and aircraft safety. Furthermore, Periyar Selvam et al. [19] reported that maintenance costs account for approximately 10% to 20% of total aircraft operating costs.

Rashid et al. [20] and Saleh et al. [21] conducted a study to investigate safety issues related to maintenance. Analysis of human factors in both studies shows that most of the factors influencing the maintenance event are related to improper component inspection and installation procedures, along with temporary factors that are profoundly come from the organizational and management levels.

According to a study conducted by [22], based on International Civil Aviation Organization (ICAO) official reports, approximately 0.9 percent of aircraft incidents involved maintenance. It also stated that maintenance accidents were 20% more likely to result in one or more deaths than all official ICAO accidents (14.7 percent). Nonetheless, it was discovered that the number of accidents caused by maintenance errors per year decreased over the study period, and the statistic rate decreased from 5% per year to 2% per year. The findings revealed that aircraft with maintenance contributions were typically between the ages of 10 and 20 years, with aircraft over 18 years old being more likely to result in fleet loss and aircraft over 34 years old being more likely to result in death. Several studies [13, 24] discovered that the primary causes of technical failures were inadequate repair procedures, irresponsibility, and incorrect installations, which could be caused by workforce skill sets.

Geibel et al. [24] presented the study’s findings by looking into the issues that cause errors in air service maintenance. According to the paper, the top five categories of maintenance errors are maintenance technician qualifications, maintenance inspections, spare parts installation, contract maintenance problems, and log documentation. This investigation included 1,000 NASA US incident reports for aircraft maintenance issues discovered between July 1997 and August 2006. According to the study, more than 53% of the negative results analysed in data collected from their aviation safety system. The main contributor factors were skill-based errors such as slipping errors, impairment, and perception, followed by routine violations (15%) and errors in making results (9%).

Reynolds et al. [25] conducted a study to investigate the effects of training on human factors for maintenance work forces in the European Union (EU) and the United States in order to reduce maintenance incidents (US). Data comparisons on vocational training prior to the implementation of human factor training from 1991 to 1998 and after the implementation of vocational training from 2000 to 2006 were compiled for the study. According to the study, after the introduction of trained human factors, aircraft mechanical accidents in the EU decreased from 33% to 22%, while mechanical accidents increased in the US. Ng and Li [26] conducted another study that provided a theoretical support concept to aid in the analysis of aircraft maintenance incidents. The study looked into the root cause of 109 aircraft maintenance incidents using data from several airlines. According to the findings, more than 60% of aircraft maintenance activities can be based on Rasmussen SRC framework regulations, while nearly 50% of accidents can be supported by standard aircraft types. Technical personnel’s errors or work factors

Schmidt et al. [27] and Illankoon et al. [14] described the Human Factor Analysis and Classification System (HFACS) for analysing maintenance accidents in military aviation in their research. According to [27], the influence of human factors on aviation maintenance accidents is the circumstances of managers, maintenance, and latent work, which can influence employee performance. This conclusion was made after analysing data from 470 maintenance-related accidents obtained from the Naval Security Center Information Management System between 1990 and 1997. Later research [14] reported their study on data from a fleet of fighter jets to look for maintenance irregularities reported over 38 months in January 2013, taxonomy to locate and identify hidden causes. Similarly, work attention, memory errors, work process deficiencies, and documentation are identified as important causal workforce factors in the study. This study also revealed how IR4.0 interventions can aid in the reduction of maintenance disorders while capturing hidden causal factors.

Many of the previous works discussed critical issues concerning human factor and skill mismatch errors in aircraft maintenance. This paper describes a survey of document analytics using document mining on additional published articles related to these issues by evaluating the gist content of related topics in a document dataset. As a result, it is expected that this analysis will identify human factors error, particularly skills mismatch among aircraft maintenance technicians in the discussions, as a common talent management issue in the aircraft maintenance field.

2.3 Impact of IR4.0 on aircraft maintenance

Evidently, the first industrial revolution saw a significant increase in the level of industrialization, with the transition from manual and animal power to steam power [28, 29] highlighted that the second industrial revolution was entirely driven by mass production, electricity, assembly lines, industrial chemistry and metallurgy automobiles, aircraft, radio, and the massive start of construction works and the development of more powerful engines. In the third industrial revolution, massive computers and the information and communication technology (ICT) revolution were introduced into production automation of processes, ushering in the digital age [30].

IR4.0 brings in Industry 4.0 by preparing the manufacturing sector to embrace digital transformation through the incorporation of sensing devices in virtually all manufacturing parts, products, and equipment [63]. Organizations must evaluate the impact of Industry 4.0 technology to sustainability and economy benefits [68–70]. The implementation of a ubiquitous system promotes the ability to analyse digital data and physical equipment, allowing every global industrial sector to advance much more quickly [31]. This implied Industry 4.0 is part of an interconnected world that has been altered by the ICT revolution for example the technologies merger such as cyber-physical systems (CPS), the Internet of things (IoT), Big Data and cloud manufacturing [32, 33]. [61, 66] also formalized interpretation Artificial intelligence (AI), Big Data, IoT, the Internet of services (IoS), CPS, and smart manufacturing are all listed as critical components of IR4.0. As a result, this gyration is the culmination of all previous revolutions’ inventions. According to the literature, this revolution will convert and have an irreversible impact on the built environment industry as profoundly and irreversibly as each of its three predecessors, and more rapidly as any of them. As a result, in order to handle advanced aircraft systems and manage data analytics for predictive maintenance [67], such as deep data and probably more performance analytics, aircraft maintenance technicians are anticipated to obtain more 21st-century skills [53].

In the aviation industry, one example of IR4.0 adaptation is the development of predictive maintenance systems. Predictive maintenance includes the role of sensors and IoT in maintaining and monitoring aircraft components to be replaced before visible defects emerge [64]. This approach teaches a valuable lesson in how typical aircraft maintenance work practises can be translated by the smart factory [34]. Similarly, in the manufacturing industry, components of smart factories are introduced as part of engineering science integrations. The distributed manufacturing branches or alliances are linked by leveraging technological advances such as big data analytics and remote monitoring to assist factories in increasing productivity, creating a safe environment, and maintaining safe operations [34]. Today’s technical advancements in the aviation sector have impacted aircraft maintenance professionals, making it necessary for them to become compatible with more sophisticated aircraft systems. In order for maintenance technicians to attain this compliance, technician training in organisations and institutions that provide maintenance training must be tailored to today’s circumstances [51]. Maintenance-related delays may be avoided significantly by having well-trained and experienced aeroplane technicians with experience in specialised equipment. Graduates of educational institutions did not have enough training to keep up with the evolving technologies in the aviation maintenance environment [52].

2.4 Document mining

Document mining is one of techniques in Data Mining (DM) [35, 36]. It aims to discover knowledge in a stack of documents that may have comparable discussions and arguments by determining term frequencies and term connections in the documents. DM is famously known as a process of extracting nontrivial, useful and meaningful insights in databases or Knowledge Discovery in Databases (KDD). KDD has an iterative multi-stage of 5 main processes as illustrated in Fig. 1. The stages are:

Selection – raw data is carefully selected for mining process.

OPEN IN VIEWER

Fig. 1

KDD stages [37].

Basically, document analytics adapt the KDD processes. A document is regularly treated as an individual information item and an assortment of it is considered as a document dataset. A proper document collection is importance for formal and informal reader to access information as knowledge and information sharing [38]. ICT allows many documents to be collected as data files which have been created or converted into digital papers such as articles, contracts, reports and news. Thus, for faster information sharing, document mining can be applied to process digital documents. Document mining consists of several techniques of information extraction, natural language processing (NLP), information retrieval and text mining. The key use of document mining is to explicitly identify the term association and similarity in a document dataset [39, 40].

Text mining (TM) is one of data mining methods for extracting insights using text data. TM literally a text analytics technique that incorporates NLP and DM techniques to convert the free (unstructured) text in documents and databases into standardized and structured data patterns. NLP is used during the preprocessing stage to clean, convert, and standardize text data (text terms) formatting.

In TM, cleaned terms undergo a ranking process using Term Frequency-Inverse Document Frequency (TF-IDF) to assign term count and term weighting for each term that appears within set of documents or corpus. Thus, it produces a list of terms ranking based on TF-IDF. Then, Term Document Matrix (TDM), a two– dimensional matrix or table, is constructed with rows represent terms and columns contain document ids in a corpus. The purpose of generating TDM is to record calculated terms’ frequency counts of terms across documents in a corpus [41].

Figure 2 shows a general process of building TDM from a corpus using TF-IDF equation. TDM contains list term entries of (i,j) that represents the frequency count of the text term i which appears in document j. The range of a term’s frequency count is in between 0 and n. Hence, rows in TDM records the terms similarities between one document to another document in the same corpus.

OPEN IN VIEWER

Fig. 2

Term document matrix generation.

TF-IDF is a most important method to compute term’s frequency counts and weights across documents in a corpus that frequently utilized in information retrieval and TM. It is supposed to measure how significant a term is to a document in a corpus. Therefore, TF is the term frequency for the terms appear in a document while IDF is applied to measure how significant of a term for the collection of documents or corpus [41, 42]. Each term has its corresponding TF and IDF score and TF^*IDF weight of each term using equation shown in Equation 1: w i , j = tf i , j × log ( N df i )

tf_i,j is total of frequencies of i in j.

The w_i,j value of a term increases correspondingly to the number terms appear in the documents. It is used to examine how significant the term is all through documents in a corpus.

Document mining may appear technical to non-data science experts, but there are many practical text or document analytics tools on the market, such as SAS, WordStat, Voyant Tools, RapidMiner, and many more. Some of these are cleverly packaged, powerful, and ready for non-technical users to use in developing their own text or document analytics tasks.

In this experiment, we use a document analytics tool to assess a set of online articles that discuss aircraft maintenance and skill mismatch issues. We integrate NLP and TF-IDF as an absolutely combination method.

Human resource management approaches such as placing individuals in certain roles without the necessary skills or education might hurt a company’s competitiveness yet declaring so does not create an image that can help the company especially in aircraft maintenance. [45] conducted a study on Air Asia aircraft maintenance technicians agreed that skill mismatch affects their work performance. 50% of the respondents (technicians) agree that some of the tasks assigned are mismatched between their capabilities because of limitation of knowledge and skills with the new difficult and unfamiliar task with limited time given. [46] also studied that majority aircraft incidents in Nigeria between 2006 and 2019 are affected by maintenance factors with the highest contribution to these accidents are “operator and regulatory oversight”, “inadequate inspection” and “failure to follow procedures”. These are in line with the Figs. 7 and 8 which illustrate ‘employ’, ‘job’, ‘requir’, ‘skill’ and ‘mismatch’. [47] examined the skills mismatch dimension, which looked at the gap between education and the employment, including between employee skills and company demands. The first type of mismatch is termed educational mismatch, and it refers to situations in which individual’s education is inconsistent with the requirements of the job. The second educational mismatch is undereducation, which occurs when an employee’s education falls short of what the organisation requires.

Work-based learning, or in-house specialised training that targets the specific needs of the employer, is frequently used to overcome the skills mismatch [48, 49]. Furthermore, tailored training is considered as a way to meet the many “levels” of skill requirements that businesses require. The gap between the needs of the aviation workforce and the skills of the technician prospect can be bridged through tailored training. Airport Region Council in Belgium oversees a project on re-designing of work processes and conducting a work-based learning for mismatch capability aviation worker [50]. This is consistent with the findings in Fig. 9 that depicts terms ‘skill’, ‘educ’, ‘employ’, ‘use’ and ‘studi’ extracted from the document set.

In recent years, the aviation industry has experienced a growing interest for reskilling and upskilling [6]. The aircraft maintenance workforce will endure skill development programmes, as outlined and proposed below:

Personnel upskilling [2, 62]: Management will be expected to upskill its employees, either domestically or through external training institutions. A first-line technician, for example, will be required to manually fit a part using a robot or other IoT devices. He or she should learn to use the new tools effectively. Internationally, the growth of skills for the continuous maintenance knowledge base falls behind that of manufacturing technologies and systems.