ASTRYX Marine Intelligence

Commercial fishing operators often rely on experience, fragmented environmental information and local sonar readings when deciding where to search for fish. This can lead to unnecessary search time, higher fuel consumption and unsuccessful trips. ASTRYX is an AI powered marine intelligence platform that transforms satellite, weather and oceanographic data into practical recommendations. It analyses sea surface temperature, chlorophyll concentration, turbidity, wind and water currents to identify and rank zones with potentially favourable fishing conditions. The platform displays priority zones on an interactive map and supports trip planning, weather assessment, route selection and performance reporting. Operators can record trip routes, fuel consumption and catch results, allowing ASTRYX to evaluate predictions and improve future recommendations. ASTRYX narrows large search areas before a vessel begins local verification with its existing sonar. An optional underwater drone is also being developed for targeted environmental and visual verification. Our initial customers are small and medium commercial fishing operators in the Caspian and Black Sea regions. Revenue will come from paid pilots, monthly subscriptions per vessel, fleet plans and customised enterprise contracts. ASTRYX currently has a functional web prototype, an initial underwater drone platform and recognition through Azerbaijan’s First National Space Hackathon and startup incubation programmes. Our next milestone is to conduct field pilots and measure improvements in search time, fuel use and operational efficiency.

Номинация

AI-стартап года

Компания

Astryx

Проект

AI supported platform to detect fishing zones for the fishing companies

Сфера

AgriTech / сельское хозяйство

Стадия

Идея / прототип

AI-компонент

ASTRYX uses an explainable geospatial suitability model rather than a generative AI system. The platform divides the selected marine area into map cells and processes environmental variables including sea surface temperature, chlorophyll concentration, turbidity, wind and water currents. The data is cleaned, normalised and aligned by location and time. For each cell, the system compares current conditions with species specific environmental preferences. Suitability functions and weighted parameters are combined into a Fish Potential Score, which ranks areas according to their probability of having favourable fishing conditions. The platform then displays priority zones and the environmental factors contributing to each score. The current prototype uses an explainable scoring approach so its recommendations can be understood and adjusted. The next machine learning stage will use real trip outcomes, including route, environmental conditions, catch quantity, search time and fuel consumption, as labelled data. A supervised classification and ranking model will learn local relationships and improve recommendations for particular species and regions. Training and evaluation data will be separated by trip, location and time to prevent data leakage. Performance will be measured through precision and recall for productive zones, ranking quality and comparison with actual catch outcomes. ASTRYX presents predictions as decision support, not confirmed fish locations. Existing vessel sonar and the optional underwater drone can provide local verification and additional feedback data.

Влияние / результаты

ASTRYX has developed one functional web prototype that combines four principal environmental layers: sea surface temperature, chlorophyll concentration, turbidity and wind. The prototype includes an interactive map, area selection, environmental visualisation, priority-zone scoring and reporting functions. An initial underwater drone platform has also been assembled for the future verification component, although full sensor integration and field validation are still in progress. The project achieved third place in Azerbaijan’s First National Space Hackathon, continued its development through a startup incubation programme and was presented to visitors, ecosystem representatives and investors at Baku ID 2026. ASTRYX has not yet reported unverified fuel savings or prediction accuracy as completed results. The next milestone is a controlled pilot with commercial fishing vessels. Baseline and post-use measurements will include search duration, distance travelled while searching, fuel consumption, successful versus unsuccessful trips and catch per unit of effort. The initial pilot objective is to determine whether using ASTRYX produces measurable improvement compared with the operator’s normal decision-making process. These results will be used to validate the model, calculate customer ROI and establish evidence-based commercial pricing.

Инновация

Most existing marine services provide raw satellite maps, weather information or local sonar readings as separate tools. ASTRYX combines these sources into one operational workflow designed specifically for commercial fishing decisions. Its primary difference is the connection between large-area pre-trip analysis and local verification. Satellite and oceanographic data help operators identify priority zones before travelling, while existing vessel sonar or the optional underwater drone can verify conditions within the selected area. ASTRYX also creates a feedback loop between predictions and real operational results. Trip routes, environmental conditions, fuel use and catch outcomes can be used to evaluate previous recommendations and progressively localise the model. Unlike platforms that only visualise data, ASTRYX is designed to provide ranked zones, trip-planning support and performance reports. For Azerbaijan, it introduces a locally developed combination of Earth-observation technology, explainable AI and marine operations focused initially on the Caspian and Black Sea regions.

Связь с Азербайджаном

ASTRYX was founded and developed by an Azerbaijani team to address challenges faced by fishing operators in the Caspian region. The project builds on Azerbaijan’s growing AI, space technology and Blue Economy ecosystem by converting satellite and oceanographic data into practical commercial applications. ASTRYX originated through Azerbaijan’s First National Space Hackathon, where it achieved third place, and continued through local startup incubation activities. The project was also presented at Baku ID 2026. Its first validation market is Azerbaijan. The team plans to work with local licensed fishing operators, marine researchers and relevant public institutions to collect operational feedback and conduct field pilots. The resulting experience can support expansion into other Caspian and Black Sea markets while keeping product development and AI.

Команда

1–5 человек

Бизнес-модель

B2B SaaS / лицензирование

Дата запуска

2026-08

← Все номинанты