Le projet de recherche AmphibiAR étudie le potentiel d'un système LiDAR et SONAR intégré pour la surveillance et le contrôle des infrastructures portuaires. L'intégration du LiDAR, du SONAR et des capteurs de navigation sur un navire forme un amphibien artificiel qui peut voir sous et au-dessus de la surface de l'eau. La méthode que nous proposons permet de garantir que notre système assurera la qualité des données dans les environnements dépourvus de signaux GNSS à proximité d’infrastructures d’envergures. La solution développée intègre un scanneur LiDAR Velodyne VLP-16 et un système de navigation GNSS/IMU. L’environnement ROS1 est utilisé pour communiquer avec les différents capteurs, enregistrer leurs données et les acheminer vers des algorithmes SLAM2 qui ont été adaptés pour faire la correction des trajectoires erronées. Avant d’appliquer cette solution sur des données réelles, un simulateur a été conçu pour générer des données LiDAR dans le format VLP-16 d’une trajectoire de contrôle considérée sans erreur et d’une trajectoire erronée. L’application d’un algorithme de correction basée sur la technique ICP3 a démontré son efficacité à corriger la trajectoire erronée dans un environnement simulé. L’erreur planimétrique, originalement établi à 50 cm, a été réduite à un ordre inférieur à 10 cm à la suite de l’application de cet algorithme. La prochaine étape consistera à acquérir les données réelles avec un système LiDAR-SONAR mobile monté sur différentes plateformes marines afin de valider la qualité de la trajectoire estimée selon l’approche présentée ici.

Benthic ecosystems, which include the diverse assemblage of organisms inhabiting the ocean floor, play a crucial role in maintaining the health and functioning of aquatic environments worldwide. Monitoring and understanding these ecosystems is paramount for effective conservation and management efforts. In recent years, the adoption of automatic benthic inventories, powered by technological advancements in underwater robotics, remote sensing, and artificial intelligence, has revolutionized our ability to study and assess benthic habitats. Automatic benthic inventories offer a significant advantage in terms of efficiency. Traditional manual sampling methods are time-consuming, labour intensive, and often require a substantial financial investment. In contrast, automated systems equipped with high-resolution cameras, sensors, and autonomous navigation capabilities can cover large areas in a fraction of the time. This improved efficiency allows researchers to gather data more frequently. Additionally, the accuracy of automatic benthic inventories is a notable advantage. These systems can capture detailed images and collect precise data with minimal human interference, reducing the potential for observer bias and errors associated with manual sampling. High-resolution imagery and advanced machine learning algorithms further enhance the accuracy of species identification and habitat characterization, leading to more reliable ecological assessments. Furthermore, automatic benthic inventories contribute to the safety of researchers and marine environments. Operating in challenging and often remote underwater environments, these systems reduce the risks associated with diver-based sampling, such as decompression sickness and physical injuries. Moreover, their non-invasive nature ensures minimal disturbance to fragile benthic ecosystems, thereby preserving their natural state for long-term research and conservation efforts. We will showcase the development of a state-of-the-art automatic marine inventory pipeline built using marine robotics, photogrammetry, and artificial intelligence. A case-study using data acquired on the north-shore of Quebec will be used to demonstrate the efficiency and capabilities of the system, going from automated acquisition, to the generation of actionable hydrospatial intelligence products about benthic organisms.

As a user of a sub-bottom system, either for sediment imaging or for detection of buried objects, your main concern may be, will my sub-bottom profiler system be able to detect sub-bottom layers deep enough for my purpose, or will buried features be detected. In Kongsberg Discovery we offer a wide range of sub-bottom systems. The different systems have different operating frequency ranges, different transducer configurations and thus beam characteristics, different physical sizes and configurations, and different price. It is the physical characteristics of the system that determines the performance, but as long as the sediment conditions by default are unknown, we will be careful trying to quantify how deep the signal will reach. We would rather that you as a user understand the theoretical and practical meaning of the system parameters, and by that be able to plan for and explain your sub-bottom data. This presentation will start off with the sonar equation, and further in a non-expert manner explain and show examples of how system parameters such as beam width and frequency range affects sub-bottom performance. In addition, the presentation will provide some examples of how external factors such as sediment characteristics and external noise comes in as unknowns into the equation and affect the sediment imaging. Some recommendations will be given how to configure your system to best compensate for pitfalls. The presentation will use examples from Kongsberg’s own sub-bottom profiler systems to elaborate and demonstrate above mentioned performance dependencies. The presentation will include brief product overviews of the Kongsberg sub-bottom profiler systems SBP29, TOPAS, EA, as well as the newest concept EM SBP. EM SBP provides sub-bottom functionality with the EM 304 or EM 124 multibeam echo sounder systems without need for additional hardware.

In Spring 2023, Whiteout Solutions embarked on an ambitious and ecologically significant project to map over 14,000 acres of Connecticut’s protected coastal marshlands with UAS and helicopter based topobathymetric LiDAR. The project objectives were to establish a baseline elevation dataset for monitoring and validation habitat restoration techniques crucial for maintaining biodiversity in these sensitive areas. Bird species such as saltmarsh sparrow are declining, and declines of other species are expected as marshes flood more frequently and marsh vegetation changes. Habitat transitions occur along a very shallow elevation gradient in tidal marshes, and because sea level rise is predicted to occur at rates of ~3 mm per year in New England, elevation data needs to be of high precision to understand how future conditions may alter the plant communities of tidal marshes. At the heart of the data collection process was the Riegl VQ840 GL topobathymetric LiDAR scanner. This scanner provided detailed topobathymetric data, 200 elevation points per square meter of the marsh, intertidal zones, and channels. Complementing the LiDAR scanner was a hyperspectral line scanning sensor capturing detailed spectral information, enabling the identification of various marshland vegetation types and conditions. The final data products will include high resolution terrain models that showcase microtopologies and vegetation land cover maps to be used for habitat assessments. These will serve as invaluable tools for environmental scientists, conservationists, and policy makers, aiding in ongoing efforts to monitor, preserve, and restore Connecticut’s precious coastal marshlands. From Connecticut’s coastal marshlands to the shoreline of the St. Lawrence River we’re seeing the first order impacts of climate change. Continued innovations in remote sensing and unmanned aerial systems, are giving us the ability to conduct timelier and high resolution surveys and soon may translate to highly impactful solutions for resiliency.

As ocean temperatures and sea levels continue to increase and weather storms become more severe and frequent, benthic habitats such as coral reefs become more vulnerable to deadly conditions such as mass bleaching events and infectious diseases. With these ever-changing conditions, it becomes imperative that monitoring efforts are made to ensure longevity of the world’s coral reefs. Current coral reef monitoring techniques require significant manpower to collect in-situ data, such as fixed site surveys using photography and visual counts. The Ice, Cloud, and land Elevation Satellite-2 (ICESat-2), while initially intended for collecting data regarding changes in the cryosphere, utilizes a green laser thus opening the door for an abundance of oceanic and bathymetric applications. ICESat-2 is currently an underutilized data source for ocean-related purposes, despite its high resolution and frequency, as well as the economical alternatives it potentially offers the remote sensing community. Using ICESat-2 as the primary data source, in this study machine learning methods are used in the detection of coral reefs located around Heron Island, Australia. Classic ICESat-2 variables such as date, depth and geographical location are used in conjunction with algorithmically extracted features of the seafloor such as a window-based pseudo-rugosity measurement. Binary logistic regression results are promising, motivating a comparison with convolutional neural network results. Both machine learning models show that the addition of Sentinel-2 satellite derived bathymetry values increase accuracies of coral detection. This research suggests ICESat-2 to be a useful data source in future coral reef monitoring methodologies. Ongoing work examines the value of automated reef identification in developing monitoring methodologies, as well as the value of other information that can be extracted from ICESat-2 data alone. Future steps will explore the applicability of these results to other types of reefs or benthic habitats.