International Scientific Journals
of Scientific Technical Union of Mechanical Engineering "Industry 4.0"

Modern geospatial systems increasingly require search that is both where-aware and meaning-aware. Traditional spatial indexes (e.g., R-tree, Quad/Oct-tree, S2/Geohash) excel at geometric predicates and topological filtering, yet fall short when users ask semantic questions (“ports similar to Rotterdam,” “neighborhoods with transit-oriented development like X”). In parallel, embedding models for text and imagery enable powerful semantic retrieval but typically ignore spatial topology, containment, and scale.
This paper introduces a hybrid spatial–vector search architecture that unifies spatial predicates with embedding similarity for GIS-scale data. The proposed approach involves: (i) a two-stage retrieval process that initially prunes candidates using spatial cells (such as R-tree or S2 indexing) before ranking results with approximate nearest neighbour (ANN) search over embeddings (for example, HNSW or IVF methods); (ii) cell-aware vector indexes that co-partition embeddings according to space-filling curves, thereby reducing cross-cell probes; (iii) a cost-based query planner designed to jointly optimise spatial selectivity and vector recall; and (iv) a multi-modal Retrieval-Augmented Generation (RAG) layer, which integrates map features, textual data, and remote-sensing image embeddings to produce grounded responses. Evaluation is conducted on public geo-text and satellite imagery datasets, with results reported on latency/recall trade-offs, spatial bias effects, and robustness across heterogeneous scales and coordinate reference systems.
Results demonstrate that hybrid indexing delivers more than tenfold lower latency at fixed recall compared to vector-only baselines for spatially selective queries, while maintaining geometric correctness through predicate pushdown. Integration pathways with mainstream GIS and spatial SQL systems (such as PostGIS combined with pgvector) are explored, and ongoing challenges are identified in areas including geodesic distance metrics, CRS normalization, privacy, and reproducible benchmarking. These findings provide a practical blueprint for AI-empowered geospatial search that addresses both the spatial characteristics of locations and the semantic aspects of meaning.

The integration of the Semantic Kernel with OpenAI presents a novel framework for developing agentic artificial intelligence (AI) solutions for autonomous environmental control in smart homes. This approach leverages Semantic Kernel’s capabilities in natural language understanding, contextual reasoning, and task orchestration, combined with OpenAI’s advanced generative AI models. Together, these technologies enable the creation of intelligent agents capable of interpreting complex user commands, understanding contextual nuances, and autonomously managing dynamic environmental conditions within smart home ecosystems.
The solution meets the challenges of making real-time decisions and providing personalized user experiences in smart homes. Semantic Kernel allows the design of flexible AI agents that manage memory, interact with external APIs, and execute tasks efficiently. When combined with OpenAI, these agents acquire superior language processing and conversational skills, facilitating seamless interactions with users and other smart home devices. This leads to a smart home ecosystem where AI optimizes lighting, temperature, quality, and energy use based on user preferences and context.
A significant feature of this framework is its capacity to function autonomously while adjusting to user input and evolving scenarios. Semantic Kernel’s contextual memory facilitates tailored interactions by retaining user preferences and previous actions, enabling AI agents to modify their responses accordingly. For instance, the system can learn and adapt to a user’s preferred lighting and temperature settings throughout the day or respond to changes in weather or energy availability.
The framework includes design principles and implementation strategies for integrating Semantic Kernel with OpenAI in smart home environments. Practical examples show how these technologies can turn traditional smart homes into fully autonomous systems. Demonstrations cover scenarios like coordinating multiple devices for energy efficiency, adapting environmental control based on user activity, and AI-driven emergency responses for safety.

Data analysis, particularly predictive analysis, has seen significant development with the emergence of modern LLMs,
especially the latest GPT models. Nowadays, Generative AI is a turning point in modern industry. Many new LLMs have been available
during the last several years, but an essential case is how these models are changing: what is the direction of evolution, and what can we
expect from Gen AI shortly?
This study analyses different aspects of the evolution of predictive analysis systems using different LLMs: GPT 3.5 turbo, GPT-4, GPT-
4o, GPT-4o1 OpenAI GPT o1.
This report includes comparative studies of predictive analysis for construction structures conducted using a comparative reference
framework with different GPT models, reaching GPT-4o and GPT o1. The accuracy of the analysis on identical cases is compared. The
study also compares the cost and performance of the different models. Current research will be useful for scientists, researchers, industry
engineers, and businesses to estimate the cost and effectiveness of the GenAI solutions that they are going to implement and choose the right
LLM for their cases and

Nowadays, many structures should be inspected, analyzed, and repaired. This is a complex and expensive process that also includes predictive analytics to prevent possible construction failures.
One of the most used predictive analytics applications involves extracting necessary metadata from images and videos to evaluate the condition of real-world systems and recommend measures to sustain these systems. Image analysis is not a new concept – many solutions have been used for several decades.
The current paper mainly focuses on OpenAI-based capabilities to implement Image Analysis and Cognitive Digital Twins and proposes faster, cheaper implementation and more adaptive approaches to offering predictive analysis for constructions.
ChatGPT (Chat Generative Pre-Trained Transformer) is one of the trending technologies in modern Artificial Intelligence (AI), and experts in this area expect to have a very high impact on the industry shortly.
One of the latest versions – GPT-4 Turbo with Vision, developed by OpenAI, is a significant multimodal model (LMM) capable of interpreting images and providing text-based answers to queries regarding those images. It combines capabilities in natural language processing and visual comprehension.
The proposed approach considers using OpenAI LLM and Digital Twins for three different aspects of predictive analysis for Construction: image analysis, case decomposition, and creation of self-adaptive models to find possible trends to compromise structures and offer preventive actions. This research compares traditional methods for inspection and repair of Construction, including the time required for predictive analysis, the correctness of the proposed actions, and the cost of the methodology.

Industry 5.0 is the production model based on the concept of how people and machines can work together. Modern industry is focusing on collaboration between people, robots, and intelligent machines with the help of AI. The design and implementation of solutions, based on Industry 5.0 brings a new level with the latest achievements in Artificial Intelligence and especially the Generative AI, realized on top of the Open AI.
This paper focuses on implementing modern automation with the help of Open AI as a new concept to use AI as a SaaS, integrated with smart machines and robots, helping them to work in a human-like way. The research overviews the use cases and proposes a framework to build a smart solution based on Generative AI and cognitive robots, where integration between SaaS AI and smart machines is based on Cognitive Digital Twins. Research results also include prototypes of cognitive digital twins and its integration with intelligent machines.