Lowering Costs of Underwater Biological Surveys to Inform Offshore Renewable Energy
Using artificial intelligence to make underwater biological surveys more efficient for offshore renewable siting.
Cal Poly Corporation
Recipient
San Luis Obispo, CA
Recipient Location
17th
Senate District
30th
Assembly District
$199,478
Amount Spent
Completed
Project Status
Project Result
The project was completed in 2020 and the final report is available online. The team completed the development of the video annotation software and the machine learning portion of the project to automate the identification of target species in the video. Students annotated underwater video from the Monterey Bay Aquarium Research Institute. The technology developed for this project is a proof of concept and achieved acceptable accuracy for several species. Additional work is needed before this tool can be used for large-scale implementation of automated classification of deep-sea organisms. Computer science students gained real-world experience coding the software, while marine biology students learned to identify species in the underwater video. The approach could be extended in the future for other applications, such as marine or terrestrial birds and bats.
View Final ReportThe Issue
As California explores opportunities to develop offshore renewable energy capacity, there will be a growing need for pre-construction biological surveys and post-construction monitoring in the challenging marine environment. Underwater video is a powerful tool to facilitate such surveys, but the interpretation of the imagery is costly and time-consuming. Emerging technologies have greatly improved automated analysis of underwater video, but these technologies are not yet accurate or accessible enough for widespread adoption in the scientific community or industries that might benefit from these tools.
Project Innovation
This agreement funded a core team of scientists, students, and staff from computer science and marine biology to develop DeepSeaAnnotations.com, a free and open-source, web-based software. The team performed three main development tasks that will lead to open-source artificial intelligence classification capabilities: 1) "intelligent" video/image annotation tools to streamline annotation/classification workflows; 2) custom convolutional neural network training using an iterative training process to improve the accuracy of the prediction model; and 3) the annotation software, workflow, and tools on the cloud to provide widespread adoption and customization capabilities for the broader scientific and consulting community. Using this tool, undergraduate marine biology students interpreted 50 hours of high-resolution, benthic survey video provided by the Monterey Bay Aquarium Research Institute, resulting in more than 40,000 annotations of more than 100 classifications of deep-sea, benthic species. These data were then used to annotate new videos for five environmentally important species and assess the accuracy.
Project Benefits
This project provides advanced tools to scientists to facilitate the efficient collection of higher quality data that will provide regulators, decision makers, and the public with greater scientific certainty regarding the impact of marine renewable energy on California's marine ecological resources. Reducing the regulatory uncertainty of marine renewable energy production will provide decision makers with better information about impacts of offshore renewables as California seeks to achieve its Renewables Portfolio Standard (60% renewable electricity by 2030) and the 100% renewable and zero-carbon electricity goal established in Senate Bill 100.
Consumer Appeal
Reducing the scientific uncertainty in the assessment of marine ecosystems and potential impacts will be important to gaining stakeholder acceptance of offshore renewable energy deployment.
Affordability
This project will lower costs by automating costly and time-consuming tasks in marine biological surveys associated with the planning and permitting of offshore renewable energy facilities.
Environmental Sustainability
This project will provide regulatory agencies, decision makers, and stakeholders with higher-quality, lower-cost data on marine ecosystems and the potential impact of offshore renewable energy technologies.
Key Project Members
Lubomir Stanchev
