Abstract:

Asynchronous online learning has become a key modality in lifelong learning, particularly in professional contexts where flexibility is essential. For more than 15 years, asynchronous hydraulic engineering courses have enabled working professionals to balance training with their job responsibilities, especially within company-sponsored training programs. However, accumulated experience and performance data reveal a recurring issue: a significant proportion of participants tend to engage with the content superficially, prioritizing course completion over a genuine understanding of the materials. This behavior often results in poor assessment outcomes and limited learning.

In order to address this challenge without compromising the flexibility of the asynchronous model, this paper proposes the integration of an artificial intelligence (AI) module aimed at fostering self-regulation and deep learning. The system analyzes learners’ interaction patterns, particularly the time spent on instructional content, and identifies cases in which navigation is excessively fast. In such situations, the AI intervenes by pausing progression and generating adaptive questions designed to assess comprehension and promote reflection.

This approach positions AI not only as a support tool, but as a mechanism for formative assessment and learner self-control. By introducing real-time, low-stakes checkpoints, the system seeks to increase learner engagement, discourage superficial behaviors, and promote deeper cognitive processes.

The paper presents the pedagogical foundations of the model and the design of the AI-based intervention system, whose implementation is planned for future editions of the course. In addition, a statistical analysis of data collected over several years is conducted to examine learners’ interaction patterns, particularly time spent on content, and their relationship with assessment performance. This analysis aims to quantitatively test the hypothesis that a significant proportion of participants engage superficially with the materials, negatively affecting their learning outcomes. Results are analyzed in terms of learner engagement, academic performance, and perceived learning quality. The conclusions highlight both the need to incorporate self-regulation mechanisms and the potential of AI-based strategies to enhance the effectiveness of asynchronous professional training without compromising its flexibility.

Keywords:

Artificial Intelligence, Asynchronous Learning, Vocational Training, Self-Regulation, Formative Assessment, Deep Learning.