October 13, 2026
In the workplace, humans are flexible; they also have the necessary abilities to simplify most complex tasks they are asked to perform. But they have their limits. Therefore, it is essential to pay special attention to the design of systems, tasks, and the conditions under which they are performed, to avoid human error.
Understanding human limitations in the workplace
Human error can be defined simply as a discrepancy between what should have happened and what actually occurred, or one that exceeds acceptable limits. As the literature points out, such discrepancies could lead to major accidents or quality defects (Torres et al. 2021). Human errors fall into three categories: lapses or oversights, errors resulting from the improper application of rules of conduct, and errors due to a lack of knowledge. Let’s take a closer look.
In certain contexts, workers perform tasks almost automatically. This can be an advantage as it is less cognitively demanding and produces accurate results. Workers have, in fact, preprogrammed their actions and carry them out much like a computer program running its course. Consider, for example, a cashier in a retail store, whose highly repetitive tasks involve scanning products, processing the customer’s payment, and sometimes bagging the merchandise. The downside of this quasi-automatic mode is that, in the event of an unexpected situation—e.g., a customer who reaches toward the cash register—the program continues to run as planned. Interrupting it takes time and can lead to inappropriate actions or omissions.
At work, thanks to their ability to simplify complex tasks, people also tend to solve problems that seem familiar using rules they’ve memorized—based on the similarity or frequency of such problems. How many of us have “jumped to conclusions” a little too quickly and made the wrong decisions lately? This is simply a matter of misclassifying situations.
Also, when faced with incomplete or incorrect information or limited resources, e.g., a lack of time, people can make poor choices. For example, you’re trying to meet a friend at a restaurant. You didn’t check a map or directions before leaving; instead, you relied on your GPS, which ultimately led you astray. You then rely on your knowledge of the city to reach your destination—an area you’re not really familiar with. Will you arrive on time? How many wrong turns will you take?
The system itself can also contribute to errors
This simplistic classification of errors underestimates those linked to the organizational context. If only it were enough to focus solely on the human… but no. Human errors also result from system design, lack of safeguards, procedures, maintenance, training, and other factors.
Beware of organizational cultures that assign blame for human errors and on scapegoating individuals. Generally speaking—and there are exceptions, of course—people try to do the right thing. If they detect an error, they will try to maintain the system’s safety by focusing only on effective actions and devoting their full attention to the steps needed to identify or correct the error.
When time is of the essence, attention can narrow, leading to a tunnel effect (selective attention). A human-centered system will have built-in mechanisms to recover from human error, including alerts sent to the worker when an error occurs (e.g., Antidote to ensure the quality of French), someone who inspects the work and brings the error to the worker’s attention (a teacher in a mentoring role), a system (digital) that semi-automatically or automatically detects and corrects errors (Tiaya et al. 2022), whether or not it includes an ethical, sustainable, and human-centered artificial intelligence algorithm (Karevan and Nadeau 2026).
Incorporating error at the onset of the design phase
It is possible to analyze the risk of human error and take action, ideally as early as during the system’s design phase. Which system functions could be affected by human error? What human actions are associated with them? What will be the effects of human errors on the system? Is it possible to estimate the probability of human errors based on expert opinion or historical data? Easier said than done…
There are several analysis techniques, ranging from traditional to systems-based approaches, including event trees, AcciMap, the Human Error Assessment and Reduction Technique (HEART), the Human Factors Analysis and Classification System (HFACS), the Systems Theoretic Accident Model and Process (STAMP), the Functional Resonance Analysis Method (FRAM), and others. It is best to work with an expert in the field.
What’s the key takeaway? Reducing human error starts, above all, with designing a system that accounts for errors from the outset; after all, to err is human. Workers must fully understand the system and be able to perform simplified tasks under ergonomic and safe conditions—whenever possible—from the design phase onward. The system must also respect the natural alignment between intentions and actions, provide feedback on performed actions, and include constraints to prevent undesirable actions.
What should we do when faced with complex tasks that cannot be simplified—or simplified to a limited extent? Indeed, many products have lifecycles spanning several decades, and many processes within our organizations are not optimized—and will not be—in terms of either effectiveness or efficiency. A redesign is sometimes too costly to be considered in the short term—e.g., redesigning an aircraft engine. In such cases, on-the-job support can help in optimizing assembly time or reducing errors (Torres et al. 2021) (Fässberg et al., 2012).
The current state of research on augmented and mixed reality deployment in complex assembly/disassembly tasks is unanimous (Zaeh et al. 2009, Keller et al. 2021, European Agency for Safety and Health at Work 2024). Nevertheless, the systems that support these new digital technologies, as well as the technologies themselves, must also be properly designed (Karevan and Nadeau 2026, Khoddammohammadi et al. 2023, Mofidi Naeini and Nadeau 2022). Several of our ongoing research projects, in both assembly/disassembly (Valdesse Eko’Ola and Nadeau 2025) and inventory management/logistics/distribution (Ngô et al. 2026) at the Applied Human Factors Engineering Laboratory—affiliated with the Products, Processes and Systems Engineering Laboratory—are presently addressing these issues.
Four important questions come to light:
- Which system functions need to be interconnected?
- What information needs to be exchanged, or what kind of assistance or support is needed?
- What technologies will enable this information to be transmitted or this support or assistance to be provided to the worker?
- How should the information be presented appropriately, or the support/assistance be provided to the worker?
The best way to reduce human error is to incorporate it from the onset of the design process for products, systems, processes, human-machine interfaces, tasks, and the conditions under which they are performed—whether the technologies are manual, automated, or digital.
For more information, please watch the webinar series from the International Ergonomics Association Resilience Engineering Technical Committee: “Systemic Methods in Complex and Digitalized Manufacturing.”
IEA Webinar #51 FRAM-STPA model: case study of an industry 4.0 plant using a digital glove: https://www.youtube.com/watch?v=lMQtC7kyi2Y
IEA Webinar #48 Using STAMP-STPA in complex manual assembly in aircraft engine manufacturing: https://www.youtube.com/watch?v=A6ViQzFVnG
References
European Agency for Safety and Health at Work (2024). Worker Exposure to Virtual and Augmented Reality and Metaverse Technologies: How Much do We Know? osha.europa.eu consulté le 8 août 2024.
Fässberg, T., Fasth, A., Hellman, F., Davidsson, A., Stahre, J. (2012). Interaction Between Complexity, Quality and Cognitive Automation. Proc. Ath CIRP Conference on Assembly Technologies and Systems, 21-22 May, USA.
Karevan, A., Nadeau, S. (2026) Integrating smart glasses and smart gloves in hybrid assembly/disassembly systems: an STPA-driven semi-automated risk management tool. Robotics and Computer-Integrated Manufacturing vol. 100.
Keller, T., Behling, M., Stockinger, C., Metternich, J., Schüt, K. (2021). Analysis of the Influence of Process Complexity and Employee Competence on the Effect of Digital Assistance in Industrial Assembly. Production Engineering, 15, 1-8.
Khodammohammadi, N., Tuyet Mai Ngô, V., Nadeau, S. (2023). Integrating smart glasses in a hybrid manufacturing system: Towards a better understanding of impacts on productivity, quality and ergonomics/human factors. CIGI Qualita MOSIM : Propulser la performance - Interconnectivité et collaboration dans un contexte d'intelligence artificielle (Trois-Rivières, QC, Canada, June 14-16, 2023) Université du Québec à Trois-Rivières.
Mofidi Naeini, A., Nadeau, S. (2022). STPA systemic approach for OHS and operational risk analysis of data glove use in 4.0 assembly. CIRP Journal of Manufacturing Science and Technology, 39, 317-331.
Ngô, V., Nadeau, S. (2026). Les méthodes d’analyses systématiques des risques, applicables au travail numérisé en équipe. Communication lors de la conférence : 46e congrès annuel AQHSST (Québec, QC, Canada, 16-17 avril 2026.
Tiaya Tedonchio, C., Nadeau, S., Boton, C., Rivest, L. (2022). Digital mock-ups as support tools for preventing risks related to energy sources in the operation stage of industrial facilities through design. Results in Engineering, 16.
Torres, Y., Nadeau, S., Landau, K. (2021). Classification and Quantification of Human Error in Manufacturing: A Case Study in Complex Manual Assembly. Applied Sciences, 11 (2), 749.
Valdesse Eko’Ola, R., Nadeau, S. (2025) Systematic Selection of Visual and Cognitive Assistance Equipment: Case of Gas Turbine Assembly in Narrow Spaces. Intelligent control and automation, 16 (3), 80-109. DOI: 10.4236/ica.2025.163005.
Zaeh, M.F., Wiesbeck, M., Stork, S., Schubö, A. (2009). A Multi-dimensional Measurement for Determining the Complexity of Manual Assembly Operations. Prod. Eng. Res. Develop., 3m 489-496.
