Formation of Adaptive Engineering Competencies: Ai Tutor as an Activator of Cognitive Synthesis
Year: 2026
Type: article
Abstract: Research objective. The integration of artificial intelligence-based systems into engineering education faces terminological ambiguity: conditional logic automation platforms are positioned by developers as genuine intelligent agents, which hinders the objective selection of educational technologies and creates the risk of substituting pedagogical development with technological imitation. The aim of this paper is to develop a criteria-based verification framework that distinguishes conditional logic automation systems from agents capable of initiating the synthesis of engineering and management heuristics that are absent from educational materials, and to describe the mechanism for forming adaptive expertise — a competency that is ahead of the requirements of Federal State Educational Standards. Materials and methods. A functional analysis of educational system architectures based on artificial intelligence, described in peer-reviewed domestic and international sources from 2024 –2026, was conducted. The theoretical and methodological foundation comprised the concepts of adaptive expertise, metacognitive regulation, and expansive learning. Three qualitative criteria with observable markers were formulated to verify the system’s ability to initiate heuristic synthesis: differentiation of error types, generation of non-canonical scenarios, and activation of heuristic synthesis. Expert calibration of the criteria was performed using the sequential evaluation method with the participation of 14 experts — practitioners in engineering and management disciplines (Kendall’s coefficient of concordance W = 0.74, p < 0.01). Demonstrative pilot testing of the methodology was carried out on a sample of master’s students in engineering disciplines (38 learners) solving a logistics problem with infrastructure constraints. Results. It was established that the vast majority of platforms labeled as “AI tutors” operate on static conditional rules and fail to initiate the construction of new algorithmic strategies. For the first time, the concept of “activation of cognitive synthesis” is introduced as an operationally measurable characteristic of an intelligent agent, tied to three observable markers. The key paradigm shift in assessment consists in transitioning from formal academic performance to recording the learner’s ability to go beyond a known algorithm. Pilot data (accuracy of error type recognition 89%, proportion of unique scenarios 63%, and proportion of master’s students who developed specialized heuristics 12.5%) are illustrative and confirm the fundamental fixability of heuristic synthesis in interaction protocols. A new paradigm of pedagogical roles is proposed: the teacher as a real-sector strategist, the intelligent system as an activator of cognitive synthesis. Conclusion. The developed toolkit resolves terminological ambiguity in the evaluation of artificial intelligence-based educational technologies and establishes a foundation for comparable verification of digital platforms in the industry. The transformation of pedagogical roles ensures the formation of adaptive expertise — a competency that is absent in the current Federal State Educational Standards but critically important for engineering activities under conditions of infrastructure uncertainty. The methodology is addressed to methodologists and teachers of engineering and management disciplines, enabling evidence-based decisions when selecting digital tools based on observable markers rather than marketing formulations. The threshold values are of a project nature and require further empirical calibration on extended samples.
Source: Open Education
Author: S. N. Gamayunov
Institution: Tver State Technical University
Language: English
Subfield: Education
Field: Social Sciences
Domain: Social Sciences
SDG: Quality education
Open Access status: diamond
OpenAlex ID: W7225914414
