Projects

Globally, fish have a market value of 150 billion EUR per year. In addition, the implementation of the European Water Framework and Habitats Directives underscores the necessity for long-term environmental monitoring across the European Union. Economically and ecologically significant fish species, such as Salmon and the critically endangered European Eel are both native to Estonia, and their life cycles require migration from marine to freshwater environments. Current academic solutions for fish monitoring are too slow and expensive, and commercial solutions with AI still rely on manual processing of thousands videos at each location. The "AquaID" project aims to develop viable systems capable of automatically detecting and counting wild fish with significantly enhanced performance. This will be achieved through the utilization of custom hardware and underwater artificial intelligence methods developed at TalTech, in collaboration with international academic and commercial partners.

This project is based on the new paradigm of "flow as information", a groundbreaking approach for underwater sensing of multiscale flows in Nature. It will lead to new, optimized devices and methods to measure, classify and explore the underwater environment when traditional methods are too expensive or simply do not work. Flow as information is inspired by aquatic animals who have evolved advanced sensory systems which combine sensing and information processing into a single framework. The proposal will advance TalTech's underwater sensing technologies from working prototypes (TRL3 to TRL5) to tests in relevant operational environments (TRL6), and support technology transfer to Estonian and international firms. These devices and methods will provide researchers, industry and authorities with new and reliable sources of flow data during extreme climate and weather events where conventional devices fail and when critical infrastructure is at risk, such as during storm surges and floods.

Data has become the most valuable resource for the automation and optimization of tasks arising both in the private and public spheres. The proposed research area/project aims at strengthening both the synergy and quality of the current research of Taltech in this area, while significantly enhancing the capabilities of Uni to cooperate with Estonian industry and public sphere by joint work, consultations, continuous and regular education. The focus of the project is on using machine learning for data science: ML, in particular deep learning, has shown the most promise in advancing the capabilities of future software systems and empowering the whole business of software development. The concrete goal is to increase the manpower and competence in machine learning, while enhancing and cooperating with the existing areas of data science like data and rule mining, data semantics and knowledge representation, natural language data queries, data integration, statistics and data management.

The project aims to redesign the current RAPID sensors to make them more fault-tolerant. This makes the testing of underwater sensor devices faster, easier, and more reliable. These sensors help researchers understand the physical conditions fish experience in hydropower plants. Currently, testing the sensors is slow and requires expert knowledge. This project will create a simple computer program that allows anyone to check the sensor's condition in just a few seconds. The results will help make future sensor systems more robust and easier to use in the field.

While there exist ways to observe the sea surface, the underwater environment requires extended research. Filling this knowledge gap is in line with EU Mission Starfish 2030. Any movement in fluid creates a pressure change. The source of such changes can be natural phenomena like waves, currents, or iceberg calving. These pressure changes also include man-made dynamics, like moving vessels or wastewater release, and can be detected and monitored using pressure sensors. These sensors are non-invasive, energy-efficient, and work submerged. The sensitivity to weak signals may suffer in presence of strong sources like waves or currents. Separation of different sources can be treated as a ”cocktail party problem.” The main purpose of this proposal is to address this challenge and contribute to underwater sea monitoring. We aim at creation of algorithms for source separation, detection and localization to work on low-cost platforms, and validate performance in real deployments.

Fish-AI is a real-time automatic video-based fish monitoring system developed by TalTech that automates reporting in accordance with EU requirements and significantly reduces labor costs. The system is compatible with various underwater cameras, detects different water conditions, and is able to adapt to them.

The project will develop a computer simulation model that will enable the assessment of fish passage conditions through hydroelectric power plant turbines. The use of simulations will significantly reduce the need for field tests, allowing for more efficient use of time and money and a broader analysis of scenarios.

The aim of this applied research is to improve the usage of radars for autonomous navigation of unmanned land vehicles in off-road scenarios. As radars use radio waves for their operation, they have several advantages over optical sensors such as cameras and lidars. For example, weather phenomena such as snow, rain, fog, etc. have much less impact on the performance of radars than optical sensors. Also, radars are not disturbed by dirt and other small debris, making them an ideal sensor for off-road navigation. However, radars do have one major drawback: the resolution of the point cloud they generate is significantly lower than that of, for example, lidars. This research will explore ways to improve the quality of data from radars using artificial intelligence and other appropriate algorithms.

Bone loading is the stress on bones during physical activity. It affects bone health and current wearable devices cannot measure it accurately. OsteoSense uses minature biosensors to capture human motion and machine learning to estimate bone loading, providing feedback and expert guidance directly on your smart phone. The system will be tested by leading experts in Estonia and professional football and racing teams in the UK, providing a world-class solution for indoor and outdoor human motion capture and bone loading reporting.

The aim of the project is to assess the fish compatibility of the newly developed pump and to compare the damage rates of the fish during passage to an existing pump in the Kudensee pumping station.

Developing, testing and validating a fully automated AI fish counting system of 20+ freshwater fish species using more than 23 Mio. videos with fish, fish swarms and changing environmental conditions.

Field study of risk of fish injury and mortality at two large hydropower plants in Norway and two in Sweden.

With a current market value of 150 billion EUR/year, fish are among the most economically valuable species on the planet. The implementation of the European Water Framework and Habitats Directives ensure that long-term environmental monitoring is in demand across the EU. Economically valuable fish species include Salmon and critically endangered European Eel (native in Estonia). They must migrate from the sea into rivers as part of their lifecycle. Existing solutions for fish monitoring are expensive, requiring manual processing of 1000s of hours of video for each location. The "ExoFish" project will provide a highly needed and valuable commercial system capable of automatically detecting and counting wild fish with significantly better performance at 50% of the cost of existing technologies, by utilizing custom hardware and underwater artificial intelligence methods developed at TalTech in collaboration with the Estonian Environmental Agency and international commercial partners.

In 2016 serious concerns on the achievement of the EU Biodiversity Strategy 2020 targets, due to the continuing loss of biodiversity and degradation of aquatic habitats, led to the urgent adoption of a new Resolution for implementing ecosystem restoration measures. Moreover, on December 2018 the EU raised to 32% the binding renewable energy target for 2030, bringing further input to hydropower development. Meeting these targets sets challenging issues for mitigating the impacts of man-made structures in rivers that fragment habitats and prevent movement and migration of aquatic organisms. The proposed ETN will train 15 ESRs in the interdisciplinary field of Ecohydraulics to find innovative solutions for freshwater fish protection and river continuity restoration in anthropogenically altered rivers. The 15 ESRs will carry out an innovative and integrated research programme within a multidisciplinary and intersectoral Network, including 8 leading European Universities, consultancy companies, public agencies and hydropower industry, encompassing experts in fish biology, river ecology, environmental fluid mechanics and hydraulic engineering. The 15 ESRs will have access to a number of laboratory and field facilities, modelling techniques, experimental practices and instrumental technologies, to expand current understanding of key fundamental fish bio-mechanical, behavioural and physiological processes, and to promote development of novel tools and management solutions in the area of freshwater fish protection. ESRs will be enrolled in specific PhD training programmes according to the rules of 6 host countries and will undertake a Network-wide training programme inclusive of research activities in at least 2 EU countries, short courses at 5 Network Schools, and a series of dissemination and public outreach actions, with the fundamental goal of forming a group of young scientists and practitioners who will play a key role in the water sector at the European scale.