SBIR/STTR Award attributes
Environmental impacts of marine hydrokinetic (MHK) installations have to be determined to determine nearfield and far field (due to energy removal) changes to the seafloor to assess whether monitoringshould be triggered. However, current ship-based tools for seafloor mapping are expensive and logistically challenging. Low-cost and user-friendly monitoring tools are sought. In Phase I project, the contractor created a remote controlled, autonomous surface vehicle with side- scan sonar to map the seafloor and biogenic habitats to assess effects of MHKs. A deep site with deployed anchors, and a shallow site with sand dollar beds were imaged. Sand dollar beds form an important coastal biogenic habitat and are sensitive to wave conditions; thus, they are excellent bio-indicators of ecologically significant levels of energy extraction. The device was demonstrated during phase I on the Oregon coast. Side-scan sonar images were obtained in locations indicative of sand dollar beds. Additional demonstrations were also conducted. The team determined that the inspection of the seabed closer to the seafloor would provide higher resolution of seafloor habitat and organisms. The introduction of machine learning could be helpful to extend the understanding of the status of such marine organisms. In the Phase II SBIR project, the team proposes to to refine and demonstrate remotely controlled low-power robotic unmanned devices to identify and map bioindicators using AI recognition and machine learning technology. During Phase II additional sensors and capabilities will be incorporated for image acquisition, analysis, and evaluation, and continue field testing. Imagery will be enhanced by incorporating a lowered towed array to avoid wave interference. The device will be able to hear, measure, image, recognize specific targets, and monitor at multiple water column depths using side-scan, 360- degree camera, reconfigurable Rigid Passive Acoustic Array, turbidity sensors, and Fish ID tracker device. 3D tracking/visualization and automated acquisition of images will enable evaluation of ecological impacts of oceanic energy extraction. Responsive sampling methods will allow the devices to automatically initiate/discontinue measurement and change sampling frequency to prescribed triggering events such as target recognition in real time. A seabed intelligence system incorporating AI of sonar and optical images on the platform will automate the analysis of where sand dollars are present and improve understanding of MHK activities impacts. Beyond MHK monitoring, the device has many commercial applications for offshore inspections and repair. This can replace dive teams and large manned ships for underwater infrastructure and integrity inspection, increasing safety and efficiency while reducing costs and manpower.
