JITA - APEIRON

Development of a System for Prediction and Optimization of Electricity Consumption in Smart Homes, Based on Artificial Intelligence

2026-06-09T16:45:07+02:00

This paper presents a machine-learning-based approach for short-term forecasting of household electricity consumption. The study aims to model temporal consumption patterns and support intelligent energy management in residential environments. Historical power consumption data were collected, cleaned, normalized and transformed into supervised learning sequences using sliding window techniques. A Long Short-Term Memory (LSTM) neural network was developed to capture time-dependent characteristics of electricity usage. The model was trained using the Adam optimization algorithm and evaluated using standard regression metrics, including Mean Absolute Error (MAE), which indicated high prediction accuracy and robustness. To ensure practical applicability, the proposed system integrates edge computing principles. Experimental results demonstrate that deep learning-based time-series forecasting can effectively predict short-term energy consumption. The proposed approach contributes to smart home energy monitoring by providing a scalable, efficient and reliable solution, and supports sustainable electricity usage through data-driven decision-making. The findings highlight the importance of integrating predictive analytics into future intelligent energy systems.

The Marquise 57 Architecture: A Diamond Fractal Geometry Approach to Robust Mobile OS (UNTLab 3327)

2026-06-09T16:53:00+02:00

This paper introduces Marquise 57, an innovative architectural framework for autonomous robotic systems, implemented and validated on the UNTLab 3327 mobile platform. Inspired by the mathematical symmetry and refractive properties of the Marquise diamond cut, the architecture employs a dual-core deconvolution strategy to eliminate logical blind spots and systemic latency.
The framework conceptualizes the robot as a digital organism, with operations divided into two cognitive layers: the Crown (Core 1), responsible for high-precision motion control via EncoMotors, and the Pavilion (Core 0), dedicated to environmental awareness through the SiLog (Sensor Information Log) protocol. At the center of this architecture is the Tower, a hardware-encrypted vault utilizing military-grade AES-256 algorithms to ensure forensic integrity of mission data.
Experimental validation over a verified distance of 1,089.57 meters demonstrates near-zero latency execution and bit-perfect forensic reconstruction of environmental events. By transforming physical stimuli into coherent informational structures, the Marquise 57 architecture establishes a robust blueprint for mission-critical cyber-physical systems operating in complex real-world environments.

Centralized Database and Distributed Communication in a Software System for High-Precision Industrial Machine Control

2026-06-09T16:59:21+02:00

High precision in industrial processes depends not only on the mechanical and control components of the system, but also on the way data related to machine operation and operator activity are organized, stored, and processed. This paper examines the application of the relational data model and distributed communication architecture in an industrial software system intended for controlling machines in which deformation, forming, and material processing are directly conditioned by precise positioning and stable process supervision. The implemented solution includes a centralized application layer developed in the Dart programming language using the Flutter framework, a distributed microcontroller layer implemented in C++, and a centralized PostgreSQL database deployed in a Docker environment on a Proxmox server. Communication between the computer and the main control node is established via a USB-UART connection, while remote actuator and sensor modules are interconnected through a CAN bus. The system supports management of operator accounts, position presets, work sequences, worker-specific tasks, event logs, and remote access via Tailscale VPN infrastructure. The paper analyzes the database structure, relationships between tables, integrity constraints, data export and import organization, as well as security mechanisms based on PIN hashing and software license protection using asymmetric cryptography. The results show that the integration of a centralized database layer and a distributed communication architecture represents a functionally and technically justified solution for this class of industrial systems.

Analysis of the expected effects of implementing ERP, GIS and DMS systems in a public enterprise

2026-06-09T17:06:54+02:00

Integrated information systems represent a key element of the digital transformation of public enterprises, enabling data integration, improvement of business process control, and an increase in transparency. This paper analyzes the effects of implementing an integrated information system in the Public Forestry Enterprise “Forests of the Republic of Srpska” JSC Sokolac, which integrates SAP ERP, a Geographic Information System (GIS), and a Document Management System (DMS).
Preliminary implementation results indicate concrete operational improvements, including the elimination of a 24-hour data synchronization delay, reduction of the field data collection cycle from 48 hours to near-real-time, instant document retrieval through the DMS, and automatic invoice generation upon mobile application synchronization. Once fully operational, the system is expected to further improve efficiency, control, transparency, and analytical management.
Special focus is placed on the identification of effects in the areas of operational efficiency, work organization, control and transparency of business operations, as well as the improvement of analytical capacities through the application of BI and OLAP technologies.
The research results indicate that the implementation of an integrated information system creates the preconditions for data centralization, standardization of business processes, and the establishment of a unified information environment. The integration of SAP ERP and GIS enables more comprehensive monitoring of the forest timber assortment production process and the improvement of analytical capacities through the application of BI tools.
A particularly significant contribution of the system is reflected in the possibility of linking planned and realized activities, thereby creating the preconditions for more efficient business monitoring and decision-making based on reliable data.
The integrated information system represents an important tool for improving efficiency and data-driven decision-making, while the achieved implementation results indicate significant potential for further development through advanced analytical methods and digital resource management.

Sentiment Analysis on Social Media Content

2026-06-09T17:12:12+02:00

Following research evaluated conventional machine learning and deep learning algorithms used for the purpose of binary text classification, in accordance with previous research demonstrating advantages in supervised learning models such as Naive Bayes, Logistic Regression, and LSTM networks. Models that were subject of implementation are: Logistic Regression, Naive Bayes, Support Vector Machine (SVM), Random Forest, and LSTM. Responses from nonprofit organizations have been cleaned, tokenized, and preprocessed implementing either TF-IDF vectorization or sequence trimming determined by the model that was chosen. The majority of the models were performed using 50,000 samples because of computational capacity limitations, whereas the LSTM was executed only with 5,000 samples. LinearSVC is implemented for the purpose of accelerating training of the SVM model, as well as Random Forest parameters optimization for algorithmic efficiency. On the other hand the LSTM model provided an embedding component and a single LSTM unit for maintaining the sequence information. The performance of the models was evaluated according to the accuracy, precision, recall, and F1 score metrics. The findings are indicating that fundamental models perform effectively and consistently, however the LSTM model demands more computational capacity to provide context for classification.

Intent-Driven Payments: A Proposed Framework for Using Large Language Models to Translate Natural Language into Structured Payment Instructions

2026-06-09T17:16:49+02:00

The modern payment infrastructure assumes well-defined inputs by users, involving specifying parameters such as the amount of the transaction, payee identity, and timing of payment execution. This mode of operation causes procedural overhead and constrains the possibility of abstraction, especially with the ongoing evolution of financial products towards greater user-friendliness. This article presents a conceptual framework for intent-driven payments, in which users formulate their financial intentions in natural language and large language models (LLMs) are used to generate corresponding structured payment workflows. While the term has been used informally in prior industry commentary, this work offers a structured architectural treatment of the paradigm. The framework described here is proposed and has not yet been empirically implemented. The intent-to-payment system presented is built as a multistep pipeline, incorporating such stages as intent extraction, entity recognition, constraint validation, and orchestration. One significant novelty of this work concerns the development of the financial intent compiler, which is designed to enforce that the outputs generated by the system are deterministic, transparent, and aligned with applicable regulatory constraints. This article touches upon a number of topics relating to the design of systems, such as latency-related problems, handling ambiguity, considerations of security, and verifications of computations made by people.

Security Analysis of the S-DES Cryptographic System

2026-06-09T17:28:15+02:00

S-DES (Simplified Data Encryption Standard) is a pedagogically oriented, reduced-complexity variant of the full DES algorithm. It operates on 8-bit plaintext blocks with a 10-bit secret key, providing a tractable environment for studying fundamental cryptanalytic methods — linear cryptanalysis, differential cryptanalysis, and combined linear-differential cryptanalysis. This paper analyzes the architectural design, subkey generation mechanism, permutation logic, and security properties of S-DES. The cipher’s vulnerability to brute-force exhaustive search and differential cryptanalytic attacks is examined in detail, and its potential use as an image encryption primitive — enhanced through chaotic key generation — is evaluated experimentally. The study concludes that, while S-DES is cryptographically inadequate for practical deployment, it constitutes a highly effective educational tool for illustrating the core principles of symmetric block cipher design.

Integrated Approaches in the Development of Intelligent Information Systems: A Comprehensive Review of Cloud, IoT, Big Data, Machine Learning, and Information Forensics Challenges

2026-06-09T17:36:24+02:00

The rapid evolution of Integrated Information Systems (IIS) has led to a complex convergence of Cloud Computing, Internet of Things (IoT), and Machine Learning (ML). While this synergy enhances computational efficiency, it introduces significant challenges in information forensics and system security. This paper explores the multidimensional security landscape of unified ecosystems, focusing on the vulnerabilities inherent in distributed resources. We analyze the necessity of “Forensic-by-Design” principles and the role of robust biometric solutions in securing e-commerce and integrated environments. Special attention is given to the impact of user interaction variability on speaker recognition performance, as well as the potential of modern IT tools in assessing and optimizing system integrity. By synthesizing recent advancements in MLOps and cloud-native architectures with empirical findings on digital literacy and security technologies, this study provides a strategic framework for developing resilient and accountable intelligent systems. The findings emphasize that technical excellence must be balanced with rigorous forensic standards to mitigate risks in increasingly dynamic, cloud-based infrastructures.

Implementation of Cowrie Honeypot System and Improvement of Log Analysis

2026-06-09T17:41:00+02:00

This paper aims to explain how honeypots work, how they are implemented, and why they have become a key aspect of cybersecurity. Honeypots are capable of doing everything from detecting new attacks never seen before in their environment to tracking programmed credit card fraud and identity theft. The paper implements the Cowrie honeypot system in a controlled environment to simulate attacks on SSH and Telnet services. Special focus is placed on the analysis of generated JSON log records, the complex structure of which makes forensic processing difficult. As a contribution to the paper, a Python helper module has been developed to convert raw log files into a readable and structured text format, thus improving the efficiency of security event analysis.