Exploring Strategies for Developing a Big Data Analytics Processing Pipeline for Aerospace Aftermarket Technical Service Organizations: A Qualitative Study

By Nathan B. Smith

Research Prospectus 

Abstract

This study addresses the strategies needed to establish a technical writing analytics application. It is envisioned that an exploratory qualitative research design approach will be followed to explore emerging technologies, identify insights, and formulate strategies. Participants will include data scientists, big data engineers, technical engineering writers working in the aviation industry, information technology managers, and commercial aviation aftermarket support experts from various maintenance, repair, and overhaul organizations, airframer companies, and military aviation intermediate maintenance departments (AIMD). The theoretical framework and constructs are intended to align with big data analytics, data mining, business intelligence (BI), machine learning (ML), and artificial intelligence (AI) with a focus on current technological advancements in big data studies (Lakshmanan, Robinson, & Munn, 2021). 

Data will be collected from approximately 15 to 20 subject matter experts using semi-structured open-ended interview questions. This study's data analysis phase will begin with organizing, labeling, and coding using a Computer-Aided Qualitative Data Application (CAQDA). Next, this textual, audio, and video data will be evaluated using Natural Language Processing (NLP) application program interfaces (API) and the Python programming language to identify themes and sentiments amongst study participants (Bird, Klein, & Loper, 2009). The primary strategies identified will include: (a) business management requirements, (b) data analytics processing pipeline, (c) the computer technology stack, (d) industry trends of big data analytics, (e) systems engineering management, and (f) other supporting environments. It is envisioned that the research findings will contribute new knowledge in the computer science field in academia and the community of technical writing practitioners while advancing the study of artificial intelligence and machine learning.

Introduction

In many technology-centric companies, under what is now being called digitalization or digital transformation, the discipline of data science, or more precisely, big data analytics, is taking a lead role in the organization and sense-making of existing and new data (Weber, 2020). The problem to be studied in the proposed dissertation is examined in aftermarket technical support in the commercial airline industry. After-market technical support encompasses the maintenance, repair, and overhaul major aircraft systems or components. 

This study will focus on aerostructures such as engine nacelles. These structures cover and provide critical support for turbofan engines. Nacelles require continuous support, especially structure repair caused by inflight and ground-based impact. As with most aircraft systems, nacelles are a critical component for an aircraft and require custom repair procedures often regulated by the FAA or European equivalent, the EASA. Structural repair procedures and associated parts lists represent a big data problem curated using antiquated, manual procedures prone to error. With the onslaught of modern big data processing pipelines, the aircraft aftermarket technical services industry can benefit from adopting artificial intelligence (AI) and machine learning (ML) (Ameisen, 2020). 

The so-called research trio consists of the study problem, the study purpose, and the research question. This concise description of the research presents a concise, integrated synopsis. The trio aligns by identifying a general area of interest (Burrington, Trio alignment: Problem, purpose, and research question, 2019). Consistent with the second legacy of Aristotle, the concept will inform the proposed research of a scholar-practitioner. A scholar-practitioner seeks to incorporate theory, personal experience, and practical work to generate scientific knowledge and promote the mission and vision of an organization and the scientific discipline (Tenkasi & Hay, 2007). In the case of this study, the area of interest lies within the discipline of computer science, with an emphasis on big data analytics. The proposed research aims to further the causes of the technical writing community, specifically within the aerospace engineering domain. 

Study Problem 

Concerning the research trio, a problem statement establishes what is known and unknown regarding the general area of interest for the specific study. The problem addressed in this proposed research study is the absence of strategic plans and best practices available to facilitate the adoption of big data analytics (BDA) into aerostructures aftermarket technical service (AATS) organizations. This topic has not been adequately considered. To date, BDA has commonly related to only large organizations. The extent of these organizations generally benefits from vast information technology and information system and personnel resources. Comparatively lower adoption in AATS organizations constitutes a significant hurdle that aerostructure repair engineers must overcome to implement BDA methodologies in their business model. To this point, only a few AATS organizations have succeeded in using only a minor portion of technology capabilities. AATS leadership and management must appreciate the benefits BDA can offer to launch the new technology strategically.

Study Purpose

The second component of the research trio is the study purpose, which indicates the objectives and goals of the project. This design science research study aims to survey the possible strategic adoption scenarios AATS organizations may choose to leverage big data analytics. These scenarios should be creative, practical, and worthwhile so that AATS organizations can use big data analytics within their business operations. This systematic investigation focuses on investigating the relevant implementation strategies by AATS organizations. The result of this study is to pinpoint the strategies that may support AATS organizations' capacity to employ this technology to drive business success, given a reasonable budget.

This study is carried out to aid AATS organizations in designing a plan of action to incorporate big data analytics into the overall business framework. Aftermarket Technical Services in the aerostructure manufacturing business sector may incorporate knowledge and data discovery, tools, and concepts that comprise the big data analytics into the business environment to maximize customer support.

Big data analytics integration in AATS organizations begins with realizing the business mission and goals, customer requirements, and the obligation to fulfill customer expectations. This information is therefore critical in suggesting an excellent approach to technological adoption for AATS organizations. The research should equip AATA organizations with possible strategies for integrating BDA. 

Research Questions

Q1 

How can ML be leveraged to improve organizational performance in an aftermarket technical services organization that curates Big Data in aircraft component maintenance procedures and associated spare parts management?

Q1a

How do aviation aftermarket support teams function? 

Q1b  

What automation techniques and tools are being used? 

Q1c

What are the concerns of the personnel in the technical writing as spare parts analysis communities in adopting AI and ML?

Q1d

What development practices are in place to adopt AI and ML?

Research Method Rationale

The research framework for this study could be either quantitative or qualitative. Research is approached through the lens of a particular worldview (Creswell & Poth, 2018). These world views may be ontology, epistemology, axiology, and researcher positionality and reflexivity. For this course, qualitative methods will be examined for suitability. The researcher's worldview will drive the selected qualitative method. The researcher may be an interpretivist, constructivist, naturalistic, or transformative. One form of qualitative analysis offers excellent value to this software-based problem in Design Science or DS. DS approach occurs in two complementary phases. First, an artifact is designed that will be used to solve a problem potentially (Dresch, Antunes, & Lacerdo, 2015). It is anticipated that the artifact, in this case, will take the form of a data pipeline. Data pipelines comprise processes that transport and transform data sources to a finished product where new and more excellent value can be realized (Hapke & Nelson, 2020). The second phase of DS represents a theoretical study of the effectiveness of the artifact in meeting the requirements met by the artifact. The second part of the research is qualitative (Burrington, Qualitative research methodology and design, 2018).

Review of the Literature Review Plan

Before embarking on data collection for this survey, a comprehensive literature review will be conducted. This section of the dissertation will give an objective and critical synopsis of peer-reviewed academic research articles typically found in online databases (Creswell J. W., 2014). As this study is focused on computer science (with emphasis is big data analytics), the following repositories will be searched for relevant articles: ProQuest (Computer science Database), ACM Digital Library, IEEE Xplore, Elsevier Science Direct, Emerald Engineering, SAGE Journals, and ProQuest Dissertations and Theses Global. Keywords and phrases will include technical writing, artificial intelligence, machine learning, deep learning, aerospace aftermarket support, Web clickstream analysis, designing data analytics pipelines, natural language processing (NLP), and content management system analytics. The Endnote software application will catalog and curate all suitable and relevant sources.

The literature review aims to build familiarity with the research topic and identify current trends in relevant research. Using critical thinking skills, a potential gap in research may bring greater focus to the research trio (problem, purpose, and research question. Themes will organize the main body of the literature review. Each article will be evaluated to identify the premise, research method, and conclusions. Author affiliation will help relate research topics to the current study. The main body of the literature review will be concluded by summarizing the key findings in general theoretical terms. Finally, all articles will be referenced using standard APA formatting (American Psychological Association, 2020).

Review of the Data Collection Plan

One of the first tasks in qualitative research is to develop a data collection plan. This plan will include a method to gather verbal data (Denizen & Lincoln, 2017). Regarding the research problem and purpose, oral data will address the research questions to determine how AI and ML could be used in the aircraft aftermarket technical services community of practice. 

According to Creswell (2014), data collection procedures set the boundaries for the research project. Further, the plan will specify how data is collected through several qualitative observations and interviews. Interviews may be highly structured or unstructured altogether. Verbal data collection, in this case, would involve designing a set of open-ended questions that could be posed to community members in face-to-face interviews. Alternatively, one could collect verbal data from existing, documented sources such as those available from the Society for Technical Communication. A third method may be based on a meta-data collection of existing academic journal articles (Morrell, 2019). 

References

Ameisen, E. (2020). Building machine learning-powered applications: Going from idea to product. Sebastopol, CA, USA: O'Reilly Media.

American Psychological Association. (2020). Publication manual of the American Psychological Association (7th ed.). Washington, DC: American Psychological Association.

Bird, S., Klein, E., & Loper, E. (2009). Natural language processing with Python. Sebastopol, CA, USA: O'Reilly Media.

Burrington, D. (2018). Qualitative research methodology and design. Doctoral Symposium (pp. 1-21). Denver: Colorado Technical University.

Burrington, D. (2019). Trio alignment: Problem, purpose, and research question. Doctoral Symposium (pp. 1-24). Denver: Colorado Technical University.

Creswell, J. W. (2014). Research design: Qualitative, quantitative, and mixed methods approach. Thousand Oaks, CA, USA: SAGE Publications.

Creswell, J. W., & Poth, C. N. (2018). Qualitative inquiry and research design (4th ed.). Thousand Oaks, CA, USA: SAGE Publications, Inc.

Denizen, N., & Lincoln, Y. S. (2017). The SAGE handbook of qualitative research. Thousand Oaks, CA, USA: SAGE Publications.

Hapke, H., & Nelson, C. (2020). Building machine learning pipelines: Automating model lifecycles with TensorFlow. Sebastopol, CA, USA: O'Reilly Media.

Lakshmanan, V., Robinson, S., & Munn, M. (2021). Machine learning design patterns: Solutions to common challenges in data preparation, model building, and MLOps. Sebastopol, CA, USA: O'Reilly Media.

Morrell, T. (2019). Library: Research techniques. Doctoral Symposium (pp. 1-18). Denver: Colorado Technical University.

Tenkasi, R. V., & Hay, G. W. (2007). Following the second legacy of Aristotle. In A. B. Shani, S. A. Mohrman, W. A. Pasmore, G. Stymne, & N. Adler, Handbook of collaborative management research (pp. 49-71). Thousand Oaks, CA, USA: SAGE Publications.