Proof of Concept Programme results 2025

Project description

SustainaBite: 3D food printing for local sustainability is a pioneering initiative of the Faculty of Management in Herceg Novi that uses advanced 3D food printing technology to transform local agricultural products into high-quality, sustainable food products. The project aims to revolutionize the culinary landscape by enabling chefs to create complex and customized food designs that were previously impossible to achieve using traditional cooking methods. Using the innovative Foodini food 3D printer, the only one of its kind in Montenegro, the team is working to optimize the use of local grains, fruits, and vegetables in printed food, optimizing taste, texture, and nutritional content.

Project description

Although the number of rooftop solar installations has grown rapidly, small wind turbines in urban areas remain largely unexplored, despite existing legislation that allows for their use. The main obstacle has been the high cost and lengthy timeframe of traditional wind assessments, which typically require a year or more of measurements to obtain adequate resource estimates. This project proposes a faster and more cost-effective alternative that will enable small and medium-sized investors to assess urban wind potential without the need for multi-year measurements. To achieve this, several rooftop stations will be installed, each equipped with an ultrasonic anemometer (and a backup cup anemometer), as well as air temperature and pressure sensors. During one season of continuous data collection, these stations will record high-frequency wind measurements and meteorological data.

Project description

The project aims to develop an early-stage prototype of a modular platform for media monitoring and trend analysis, integrating various media formats, including video, text, and scanned print content, using OCR and automated speech transcription technologies. Unlike existing tools that provide fragmented data and basic retrospective keyword monitoring, this platform introduces a unified approach: the automatic generation of concise thematic reports through the synthesis of insights from multiple media sources, with links for detailed analysis. The platform goes a step further by using multilingual NLP algorithms and temporal analysis to predict emerging narratives and changes in public sentiment before they reach their peak. This predictive layer provides a strategic advantage to companies, institutions, and media agencies, enabling them to proactively plan content, manage their brands, and prevent potential crises.

Project description

The proposed project by Fin-Ing (3D Soba) aims to develop and validate a rehabilitation glove with mechanical resistance, designed for people with partial hand dysfunction, including individuals recovering from stroke, as well as those with neuromuscular conditions such as multiple sclerosis or congenital impairments. The glove enables passive rehabilitation exercises through a spring-based resistance system, without electronics, and is suitable for use in both home and clinical settings.

Project description

Reinforced concrete is the most widely used material in construction and is characterized by high compressive strength but limited tensile strength. This limitation is addressed by incorporating ribbed reinforcing steel, which also provides other beneficial properties. For ease of transportation and handling, reinforcing steel is manufactured in standardized lengths, which in practice often do not meet the required length needed for the construction of concrete structures. One of the major challenges in construction is how to splice ribbed reinforcing steel without compromising the integrity of reinforced concrete. Therefore, the subject of this innovation project is the development of a friction coupler for splicing ribbed reinforcing steel in concrete elements. Couplers are widely used in civil and mechanical engineering structures, including the construction of high-rise buildings, towers, bridges, overpasses, tunnels, dams, and other structures.

Project description

TourBrain is a software solution that consolidates the key needs of small and medium-sized tourism agencies into a single platform. The system offers an AI editor for the automatic creation of tourism offers, a mobile application for guides and staff, a fleet module for VIP transportation, rent-a-car functionalities, and a property management module. Instead of using multiple tools for different aspects of their business—such as separate software for tours, transfers, or accommodation—agencies can now manage everything in one place, with an intuitive design and multilingual support.

Project description

Artificial Intelligence-Based Controller for Smart Reactive Power Management in Montenegro’s First Voltage-Balanced Active Distribution Network Modern trends in the development of distribution networks are focused on integrating renewable energy sources (RES) on the consumer side or in its immediate vicinity. However, fluctuations in power generation from RES integrated into so-called active distribution networks have been shown to cause voltage instability and operational problems within the grid. In addition, changing consumption patterns, driven by the increasingly widespread use of heat pumps and air-conditioning systems, significantly increase reactive power consumption, further negatively affecting voltage conditions in the network. Therefore, the integration of active distribution networks with advanced technologies that enable efficient voltage regulation is of critical importance. Within this project, the potential of applying artificial intelligence (AI) for the optimal management of various compensation devices for voltage regulation in active distribution networks will be investigated through the development of an AI controller. The proposed AI controller will enable precise monitoring and analysis of key grid parameters, with the aim of optimizing the measurement and regulation of reactive power. The ultimate goal of the project is to contribute to the development of the first voltage-balanced active distribution network in Montenegro.

Project description

The project “PAIDT PoC: AI-Driven Algorithmic Model for Trading Commodities and Indices” aims to experimentally demonstrate the technical feasibility of applying the underlying PAIDT concept. This concept, which has already proven successful in the forex/crypto domains, will be tested on new financial instruments: commodities (XAUUSD / gold) and major stock market indices (US30 / Dow Jones Industrial Average). Although PAIDT v4 currently effectively manages EUR 1.1 million, its performance in commodities and indices is unknown and presents a significant technological risk. The primary purpose of this six-month Proof of Concept (PoC) is to address the inherent technological risk associated with generalizing our sophisticated algorithms and AI models—including the generation of synthetic spread data—to new, specific market environments (commodities and stock market indices).

Project description

The VRX Montenegro project aims to develop an innovative model for valorizing Montenegro’s cultural heritage through the use of virtual reality (VR) technology. The project focuses on designing and validating a concept that enables the interpretation of historical events through immersive digital experiences, accessible to visitors regardless of physical, temporal, or logistical constraints. Kotor has been selected as the first pilot testing location. It is a city of exceptional cultural and historical value and high tourist visitation, where a significant number of sites remain inaccessible to many visitors due to infrastructure or physical barriers.

Project description

The aim of this project is to explore and implement the use of unmanned aerial vehicles (drones) in combination with a Geographic Information System (GIS) for the efficient collection, processing, and analysis of data in the territory of the village of Brezna (Municipality of Plužine). The focus is on the sustainable management of natural resources, including the identification and inventory of wild forest fruits, medicinal plants, tree species, and game, as well as the analysis of the condition of local infrastructure and facilities. Particular attention will be paid to mapping areas susceptible to risks such as floods, fires, and landslides. The collected data will be integrated into a GIS platform, creating a database for detailed analysis and decision-making support, thereby improving resource management and the efficiency of responses in crisis situations. The project will also produce a feasibility study assessing the practical applicability of this approach, its scalability, and its potential to create new jobs through the involvement of the local community. The results will serve as a model for application in other rural areas of Montenegro and the region.

Project description

The Exams4.0 project aims to transform the higher education sector through the development of a specialized large language and visual AI model (LLM/VLM) tailored to the academic context. Using high-performance computing (HPC) infrastructure, the model will be trained on a rich, multimodal dataset that includes textbooks, previous examinations, teaching materials, video recordings, and anonymized student papers. The ultimate outcome will be the development of innovative tools that enable the automated creation and grading of examinations, as well as support for curriculum planning, while ensuring objectivity and alignment with learning outcomes.

Project description

The project aims to develop an artificial intelligence-based software solution to support the analysis of pathological findings and the prevention of malignant diseases. The solution would enable the automated analysis of findings and the selection of abnormal results requiring further categorization by specialists. This would reduce the number of analyses referred to pathologists, alleviate the workload of healthcare professionals, and shorten waiting times for results. In addition to its application in the diagnostic process, the software could also play an important role in the education and training of young doctors, particularly pathology residents at the Faculty of Medicine of the University of Montenegro.

Project description

The aim of this project is to develop an innovative multi-protocol communication device – ECHO-GATE – that enables the connection of underwater sensor devices with terrestrial wireless infrastructure, addressing the key challenge of limited connectivity in underwater Internet of Things (IoT) applications. ECHO-GATE provides reliable two-way communication between underwater acoustic modems and above-water networks, such as LoRa, energy-efficient mobile technologies (LTE-M/NB-IoT), and Iridium satellite systems. This enables continuous real-time monitoring and control of underwater devices, even in remote and technically challenging marine environments.

Project description

The SAM4MI project aims to develop an innovative artificial intelligence-based acoustic monitoring system for induction motors (IM), enabling the assessment of key operating parameters—speed, torque, and current—solely on the basis of the sound signals emitted by the motor during operation. This approach provides an efficient method for monitoring IM performance without the need for direct electrical or mechanical connections to the motor, as well as for assessing the motor’s operating condition and determining the need for timely preventive maintenance. Through an experimental setup that will enable the recording of IM sounds under different load conditions and the training of neural networks, the project plans to develop a model capable of predicting the behaviour of these parameters in real time. The application of such a system could significantly advance preventive maintenance, diagnostics, and the control of electric motor drives, particularly in industrial environments where simplicity, speed, timeliness, and measurement safety are of critical importance.

Project description

Abook is an innovative digital platform dedicated to providing a premium audiobook listening experience and is also the first project of the start-up Ablest d.o.o. Through high-quality audiobooks narrated by professional actors, the project aims to transform the way people experience audiobooks. With a focus on authenticity and quality, Abook serves as a virtual bookstore where each work is brought to life through carefully selected voices and a carefully curated selection of books. Abook is more than just an audiobook listening app—it is a platform that provides literature and culture enthusiasts across the region with a unique space for listening to high-quality audiobooks and audio dramas in their native language. Accordingly, during its first year, the project plans to produce more than 50 audiobooks, including stories and fairy tales for children, popular school reading titles, as well as fiction for adults, dramas, world classics, crime novels, and works by numerous local authors, for both children and adults, alongside a carefully curated selection of literature for dedicated book lovers.

Project description

The aim of this project is to develop a software solution that automates the process of generating consolidated texts of documents, official acts, and similar materials, based on large language models (LLMs). The software will analyze amendments made to the original text, identify them, and generate an updated version containing the original text and the proposed amendments, thereby creating a so-called “consolidated text” of laws, documents, or other materials. This solution will significantly simplify the process of reviewing and publishing legal acts for government institutions, as well as for other companies and users that generate extensive documentation, such as rulebooks and procedures, and require such a tool in their work.

Project description

The project “Application of AI Technology in the Analysis of Brain Changes in Neurodegenerative Diseases and Multiple Sclerosis” aims to develop an innovative artificial intelligence (AI) model for the automated analysis of brain magnetic resonance (MR) images of patients with multiple sclerosis (MS) and neurodegenerative diseases, such as Alzheimer’s and Parkinson’s disease. Using open-source datasets as a starting point, together with state-of-the-art deep learning techniques, the model would enable longitudinal monitoring of MS progression and the degree of atrophy in specific brain regions in patients with neurodegenerative diseases through a series of MR examinations. The system would accurately quantify changes in lesion area and the volume of specific brain regions, comparing them with previous findings and reference ranges for the healthy population. In addition, the model would generate interactive visual maps clearly highlighting new or enlarged lesions, providing radiologists with an intuitive tool for the rapid identification of clinically significant changes. This integrated approach, combining temporal analysis and personalized data comparison, would represent a significant advancement over existing systems that primarily focus on static lesion detection without taking previous examinations into account.

Project description

The LOGGUARD project aims to develop a smart system for the early detection and precise geolocation of illegal logging through acoustic analysis. By using an AI model to detect characteristic chainsaw sounds, combined with GPS location data, the project plans to develop and test a functional prototype of the system. The solution will enable a rapid response by relevant authorities and more effective protection of forest resources through decentralized and energy-efficient real-time monitoring. During the project, the technical feasibility of the proposed solution will be validated under controlled conditions, creating opportunities for further system development and potential field deployment.

Project description

Flourish AI Coach is an innovation project focused on developing a next-generation digital coach that helps users develop emotional intelligence (EQ) and soft skills through real-time conversations with generative artificial intelligence. In collaboration with psychologists, the solution enables users to engage in guided, emotionally intelligent dialogues, including self-reflection, conflict simulation, and short daily questionnaires. The project will result in a laboratory prototype (TRL 4) and a feasibility study for future integration into the existing Flourish EQ platform or as a standalone spin-off solution. The innovation combines advanced AI technology with psychologically validated models, providing a safe, structured, and adaptive learning experience.