1  Introduction

Challenged to meet the demands of reduced cost, global competition, sustainability, shorter product life cycles, and product complexity, the manufacturing industry is in the midst of a digital transformation. Industry 4.0 and the underlying shift from mass production to mass customization places new demands on the workforce. Today’s manufacturing operators must manage a wider range of responsibilities and increased information flow amidst decreasing margins for error, changing methods, and new technologies. These trends show no signs of abatement (Danielsson et al., 2020). Extended Reality (XR) devices, in their various forms, are expected to aid this problem through operator training and support.

Augmented Reality (AR) systems “combine real and virtual, are interactive in real time, and are registered in 3-D” (Azuma, 1997). By realistically integrating informative and/or interactive virtual objects in our view of the world, AR aims to enhance the users’ interaction with and perception of it. Its essential affordance is the direct and natural manipulation of virtual objects in everyday surroundings. Relative to metaphorical digital interfaces, this is thought to improve the uptake of knowledge by reducing the overall cognitive load and better distributing it across multiple sensory pathways (Shelton & Hedley, 2003). AR-assisted learners demonstrate improved perception, performance, and understanding of spatial concepts, with outcomes correlated to the amount of physical engagement involved (Chen et al., 2019). As a result, AR is thought to be well-suited for task-related learning. Using untethered, hands-free devices with optical see-through head-mounted displays, AR can continuously enhance the user’s actions in the real world (Leonard & Fitzgerald, 2018). These benefits have broad industrial applications.

In manufacturing, operator support has been a common application of AR research and development since the early 1990s (Azuma & Bishop, 1994). It is also seen as a source of innovative operator training methods required to meet rapidly increasing demand for skilled labor due to high retirement rates, global expansion, and increasing specialization (Kress, 2020). Manufacturing support, training, and related applications have been identified in the areas of assembly, maintenance, operations, quality control, safety, design, visualization, logistics, and marketing (Oztemel & Gursev, 2020).

Despite great potential, the adoption of AR is slowed by technical, market, and other important social and legal obstacles (Azuma, 2019). XR technologies remain relatively immature and will face new challenges as their development moves from research labs to the shop floor. There, priorities shift from building technology to delivering solutions. No longer “proofs of concept,” these systems will be evaluated on the basis of their return on investment and other key performance indicators essential to the business case (Masood & Egger, 2019, 2020). That performance will be assessed in the context of the entire operation, where technical considerations are balanced by organizational and environmental factors. The long-term success of XR initiatives ultimately rests on how those results compare with other investment alternatives.

But AR remains a highly fragmented market, including a diverse selection of screen-based, projected, and head-mounted technologies (Kress, 2020). Studies show that the efficacy of these systems varies with the task type, technology used, application design, and other factors (Kaplan et al., 2021). Research in this area is young but accelerating. Most of it focuses on efficiency (task time) and accuracy (error count). These are relevant but incomplete measures for assessing training outcomes, where the learning rate and transfer effectiveness must also be considered (Büttner et al., 2020).

This dissertation addresses critical gaps in understanding the relative effectiveness of different AR/MR technologies for enhancing learning outcomes in manufacturing assembly training. Specifically, it explores how diverse augmented instructional methods, designed to leverage specific affordances, impact operator learning, recall, and retention in a simulated manufacturing context. The research employs an innovative affordance-based framework to evaluate and compare these technologies, providing a new theoretical lens for understanding their effectiveness.

The primary research objectives are to compare the effectiveness of various augmented instruction with traditional methods, evaluate the effect of those methods on the quality and durability of learning, investigate the role of specific affordances in those results, and explore how perceived workload, usability, and performance are related.

This study employs a rigorous, multi-phase experimental design conducted in the Tiger Motors Lean Education Lab, an environment designed to simulate modern automotive manufacturing. By combining quantitative performance metrics with qualitative user feedback, the research provides a comprehensive assessment of each instructional method’s effectiveness.

The findings of this study contribute to both theory and practice in the field of AR/MR-assisted training. Theoretically, it advances our understanding of how technological affordances translate into learning outcomes in practical settings. Methodologically, it demonstrates the value of a comprehensive, multi-phase assessment approach in capturing the full impact of these technologies on skill acquisition and retention. Practically, it offers insights to guide the implementation of AR/MR training systems in manufacturing contexts, helping to optimize both immediate performance and long-term skill development.

The remainder of this dissertation is structured as follows:

This work aims to advance the understanding of AR/MR technologies in manufacturing training, ultimately contributing to more effective workforce development in an increasingly complex industrial landscape.