Abstract
The integration of automation, artificial intelligence (AI), and robotics is, transforming the manufacturing industry, creating complex, dynamic, environments where humans, machines, and information systems interact closely. These conditions increase the likelihood of human errors, caused by cognitive challenges such as lapses in attention, poor situation, awareness (SA), flawed decision-making, distractions, communication, failures, and fatigue. Such errors result in delays, equipment damage, product defects, financial losses, and safety incidents. High SA enables, workers to better perceive, comprehend, and project information, thus, reducing human errors. Although SA has been extensively studied in, aviation, military, and healthcare domains, research focusing on SA in, manufacturing is still emerging. Given the critical role of SA in optimizing, production efficiency, ensuring safety, and minimizing errors, this study, examines the current state of research on SA in manufacturing, explores, factors that degrade SA known as SA demons, and investigates the use of, assistance systems designed to enhance SA. A structured literature, review is conducted, with findings synthesized to highlight existing, challenges and propose future research directions aimed at achieving, high SA in manufacturing environments.
Introduction
Modern manufacturing environments are characterized by complex, dynamic interactions among humans, machines, and information systems, which inherently increase the potential for human error (Landmark et al., 2019). Human errors in manufacturing frequently stem from cognitive challenges, including lapses in attention, lack of situation awareness, flawed decision-making, distractions, poor communication, and fatigue. These errors often lead to costly delays, equipment damage, product defects, financial losses, workplace accidents, and safety hazards (Jha et al., 2024). Workers with high situation awareness (SA) are better equipped to mitigate human errors due to their ability to accurately perceive, comprehend, and project information (Marquardt, 2019; Marquardt et al., 2011). According to Endsley (1990), SA is defined as the ability to perceive, comprehend, and predict future states within an environment. In manufacturing contexts, workers must be able to detect critical information, such as equipment status or deviations in processes. Missing a warning signal or a change in machine operation, for instance, can lead to undetected issues that may escalate into more severe problems. Furthermore, misinterpreting instructions or failing to understand the implications of observed changes, such as perceiving an unusual machine noise as normal and overlooking early signs of malfunction, can result in inappropriate responses (Lall et al., 2017). The concept of SA has been predominantly studied in fields such as aviation, military, and healthcare (Chi et al., 2023; Harper et al., 2023; Lee & McNair, 2024; Sugimoto et al., 2024). Recently, there has been growing interest in conducting research on SA in manufacturing, as this skill is crucial not only for optimizing production efficiency but also for ensuring workplace safety and minimizing human errors. However, there is a lack of comprehensive reviews focusing on SA in the manufacturing sector, including the associated SA demons and the methods used to develop workers’ SA skills. This review aims to examine the current state of research on SA in manufacturing, the SA demons, and investigate the use of assistance systems to enhance workers' or operators' SA. To achieve these objectives, a literature review is conducted, and the findings are presented and discussed.
Approach
The method of conducting a comprehensive review describes the review process through the following phases: databases and search strings, screening and filtering, coding, data analysis, and findings and discussion. The Preferred Reporting Items for Systematic Review and Meta-Analyses (PRISMA) 2020 statement guidelines were followed and consulted when applying the search (Page et al., 2021). Web of Science was selected as the database for its comprehensive collection of papers. Papers were selected from 2015 to 2024 and narrowed down to English journals only. The following terms were used in the search strings for titles, abstracts, and keywords: (1) the term “Situation Awareness” OR SA OR “Situational Awareness” AND Manufacturing. The term “Error,” “Safety,” “Hazard” or “Accident” were included in the search. One hundred forty-six papers were identified based on the search strings and skimmed on titles and abstracts to exclude 62. We excluded papers that are not related to manufacturing operations. The following criteria were used to narrow down the papers: (1) the article had to be conducted using a manufacturing case study, and (2) the paper had to address SA specifically, which implies that the study had to either measure factors of SA or discuss the theoretical background for SA. Lastly, 33 met all inclusion and exclusion criteria. After the papers were included, they were fully reviewed, and categories of the summary were developed based on the main idea, SA measurements, SA demons, and SA assistance systems.
Outcome
According to Adrot et al., (2021) and Stratmann and Boll, (2016), several factors undermine SA in many systems and environments. Endsley (2016) categorizes these factors into eight SA demons: attention tunneling, data overload, workload, anxiety, fatigue and stressors, complexity creep, misplaced salience, memory trap, out-of-the-loop syndrome, and erroneous mental models. Furthermore, according to the review work of Mark et al. (2021) and Zigart et al. (2023), worker assistance systems are classified into sensorial, physical, and cognitive categories. Technologies used in these systems include virtual reality, augmented reality, digital twins, and monitoring and visualization tools, etc. Sensorial systems enhance perception, while physical systems reduce physical and cognitive workload that affect situation awareness (SA). Cognitive systems support comprehension and projection. Based on our findings, attention tunneling and memory traps are frequently addressed through sensorial systems. The work of Illankoon et al. (2019), explored the lockout/tagout (LOTO) effectiveness in enhancing SA to make a safer workplace while carrying out maintenance tasks. Also, Paletta et al. (2017), Behrendt (2020), Chen et al. (2019), and Cai and Rau (2024) demonstrated the effectiveness of sensors, eye tracking, and alarm systems in enhancing attention allocation and operator SA. These technologies help prevent human errors, particularly in tasks requiring sustained focus, facilitating continuous monitoring and real-time decision-making support. Data overload and complexity creep were mitigated through cognitive assistance systems. Landmark et al. (2019), Eirinakis et al. (2021), and Zhou et al. (2018) introduced real-time monitoring, predictive analytics, tablets, smartwatches, and advanced visualization tools to reduce the volume and complexity of data presented to operators, enabling better comprehension. Furthermore, stressors, workload, fatigue, and anxiety are addressed through both physical and cognitive interventions. Illing et al. (2021) presented augmented reality (AR) as effective means to reduce physical and cognitive strain. This tools alleviate stressors, support task execution, and enhance operator SA, particularly in the second and third levels of SA. Wilson et al. (2023) noted that most SA errors in food manufactuirng industry occur at the SA level 1 and level 2 due to inability to monitor or observe data and erroreous mental model respectively. Out-of-the-loop syndrome and erroneous mental models are countered through systems that promote operator engagement and feedback. Paletta et al. (2017) and Eirinakis et al. (2021) proposed predictive SA models and continuous engagement mechanisms to ensure users remain involved in the control loop, thereby preserving their ability to respond to unexpected deviations. Florea et al. (2020), Peng and Xu (2024), Cheol Young Park and Salim (2017), and Camara Dit Pinto et al. (2024) illustrated how digital twins enhance SA, especially for anticipating and managing downgraded operational states. Predictive SA models have emerged as a central component in smart manufacturing, providing the operators with the ability to anticipate and respond to evolving production conditions. However, few studies have applied predictive SA models to enhance the third level of SA. Further research is needed in this area to promote real-time SA and improve decision-making as well as to identify SA requirements for assistance systems in manufacturing domain. Finally, SA training leveraging assistance systems is essential for preparing the future manufacturing workforce, yet this remains an underexplored area.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
