Despitethe considerable potential of large language models (LLMs) as interactive AI tutors, the systematic, large-scale evaluation of their pedagogical quality remains an open challenge. Assessment instruments currently in use are mostly based on informal criteria, or task completion is based on ad hoc criteria or binary task-completion indicators. In response, we propose a lightweight and edge-deployable classifier based on the MRBench taxonomy, which applies the principle of experts teaching and explicitly rates the tutor’s responses for guidance, actionability, and coherence. The architecture is embedded with training through rigorous Stratified Group K-Fold Cross-Validation, which avoids semantic leakage from the augmented datasets and is backed by an XGBoost model with TF-IDF vectorization. To ground statistical predictions in pedagogical theory, we add a deterministic rule layer that preserves canonical scaffolding interventions. This hybrid classifier achieves a strict out-of-fold (OOF) accuracy of 68.2% and an F1-score of 0.74. The deterministic rule layer acts as a qualitative safety net and does not statistically change the overall shift (McNemar’s test, $p=0.25$), but still helps to identify pedagogical moves that are valid, while maintaining predictive fidelity. Most importantly, single-threaded edge-simulated benchmarking of the complete end-to-end pipeline (vectorization and classification) shows a mean latency of 116.20 ms, compared to an average of 6834 ms per inference for a locally hosted 3B-parameter LLM baseline. In conclusion, this transparent and taxonomy-based framework provides a practical, real-time auditing toolkit for educators and researchers to benchmark AI tutors and observe model behavior in resource-limited deployment environments. The proposed lightweight framework ensures that robust, taxonomy-grounded tutor evaluation can be carried out without relying on costly, cloud-based LLM-as-a-judge pipelines, allowing for real-time pedagogical auditing of AI tutors in resource-limited environments.

Engineering education is based on experiential learning, but the problem is that in laboratory conditions, it is difficult to give feedback to the students in real time and personalize this feedback. The paper introduces the proposal of an innovative approach to the laboratories, called Adaptive Lab Mentor (ALM), which combines the technologies of Artificial Intelligence (AI), Internet of Things (IoT), and sensor technology to facilitate intelligent and customized laboratory setting. ALM is supported by a new real-time multimodal sensor fusion model in which a sensor-instrumented laboratory is used to record real-time electrical measurements (voltage and current) which are used in parallel with symbolic component measurements (target resistance) with a lightweight, dual-input Convolutional Neural Network (1D-CNN) running on an edge device. In this initial validation, visual context is presented as a symbolic target value, which establishes a pathway for the future integration of full computer vision. The architecture will enable monitoring of the student progress, making error diagnoses within a short time period, and provision of adaptive feedback based on information available in the context. To test this strategy, a high-fidelity model of an Ohm Laboratory was developed. LTspice was used to generate a huge amount of current and voltage time series of various circuit states. The trained model achieved 93.3% test accuracy and demonstrated that the proposed system could be applied. The ALM model, compared to the current Intelligent Tutoring Systems, is based on physical sensing and edge AI inference in real-time, as well as adaptive and safety-sensitive feedback throughout hands-on engineering demonstrations. The ALM framework serves as a blueprint for the new smart laboratory assistant.

Recent advances in educational artificial intelligence have been driven largely by large language models and other neural architectures. However, these systems continue to face challenges in factual reliability, pedagogical consistency, explainability, and governance. Neuro-symbolic artificial intelligence overcomes these limitations by combining the neural learning capabilities with symbolic reasoning to create more reliable intelligent learning systems. Yet there has been no systematic literature review that has systematically synthesized the architectural, pedagogical and governance aspects of this emerging field. We systematically searched the IEEE Xplore, Scopus, and Web of Science databases up to 10 July 2026, according to the PRISMA 2020 guidelines. A validated hybrid screening process, incorporating automated screening and required manual screening, yielded 40 studies from 3,492 records. To better reflect architecture trends, we also built an inventory of neuro-symbolic educational systems on the framework-level, which contains 92 systems. The review reveals that 60.0% of the studies included in the analysis also demonstrated architectural integration, pedagogical grounding and governance or explainability. At the framework level, 71.1% of classifiable systems adopted either unified representation or sequential integration strategies, indicating strong convergence toward a limited set of architectural paradigms. Less than half of the identified frameworks (46.7%) however, reported similar quantitative performance benefits, which restricts the ability to make robust cross-study comparisons. In general, the intersection of education and neuro-symbolic AI is heading toward a handful of well pedagogically grounded, but relatively under-developed, architectures in the area of governance and explainability. These findings provide a foundation for architectural standardization, transparent evaluation practices, and the development of trustworthy neuro-symbolic intelligent learning systems.

Introduction: Educational systems have historically adapted to incorporate technological innovations, but the success of these processes may vary greatly. Some tools become embedded in established processes, while others lead to radical transformation. In this paper, we examine the development of educational technology and recommend and critique certain technologies that facilitate existing industrial education approaches and others that require the restructuring of how educational aims and methods are defined and implemented. Purpose: This study aims to evaluate this optimization-restructuring schism in a systematic manner. We seek to show, through a century of technological transformation, that 20th-century technologies were primarily appropriated to sustain standardization and efficiency, while 21st century artificial intelligence (AI) is an unassimilable phenomenon in need of epistemological and pedagogical reworking. Method: The methodology of this research uses a Comparative Historical Analysis (CHA) approach of four critical educational technologies: the ballpoint pen, the personal computer, the internet, and artificial intelligence. This analysis is guided by a new six-dimensional architecture assessing impacts on: (1) Access & Equity, (2) Pedagogical Transformation, (3) Epistemological Foundations, (4) Student Agency & Role, (5) Teacher Role & Identity, and (6) Institutional & Systemic Effects. Each dimension was systematically coded for optimization versus restructuring impacts based on historical evidence. Results: In our analysis, we determine a distinct historical stalemate of the gravity of optimization, whereby the transformational potential of the pen, PC, and (to a smaller degree) internet, was expropriated to scale up and institutionalize the industrialized education model. Nonetheless, AI has inherent duality: it may be used as a tool of optimization (automated grading, content generation) or used as an agent of restructuring (co-creation, critical inquiry). Its capacity to generate an epistemic disruption is that it questions the authority of authorship and knowledge, renders standardized assessment somewhat redundant, and reinvents learning as a process of critical analysis and human-AI co-creation. The outcome depends on deliberate institutional choices. Conclusion: The findings suggest that the introduction of AI will have to leave the historical trends of technological assimilation behind. This will require radical policy and pedagogical shifts, such as new paradigms of assessment whose focus is on human skills, the implementation of powerful ethical AI governance systems, and the re-imagining of teacher roles as ethical mentors and facilitators of learning. The future of education relies on steering this restructuring, which is inevitable, with a commitment to equity and human-centric values.

This paper describes how we can improve Traffic Sign Recognition System to use in autonomous vehicles. The model we used was SSD-MobileNet_v2 which was very efficient. It can operate in real-time to integrate data from all the systems. This system includes obtaining images, merging sensor data and using a structured process to control the vehicle. It addresses some of the main limitations including little movements, operating in difficult lighting and performing all this in real-time on the system platform. It was able to detect items 95.9 percent accuracy in 22 frame per second. This study makes a key improvement in making autonomous vehicles safer and more intelligent.