Brian is a Software Engineer at MBARI working in the Information Engineering group of the Research and Development division. He has a bachelor’s degree in Biology from the University of Maryland, College Park and a master’s degree in Marine Science/Physical Oceanography from Moss Landing Marine Laboratories.

Brian has developed software systems supporting science at Moss Landing Marine Laboratories, the Naval Post-graduate School and at MBARI. Employed at MBARI since 1998, he has worked on numerous projects involving video and image analysis, video annotation, numerical analysis, user-interface development, and data-management systems.

Bitbucket | GitHub | JIRA

Canyon Head | MBARI Documentation | Staff Directory | Safari Books Online

Video Annotation Data Management (500055)

MBARI’s video archive contains over 30,000 hours of underwater footage collected by ROVs and AUVs — one of the largest collections of its kind in the world. Brian leads development of VARS (Video Annotation and Reference System), a suite of open-source microservices for managing and annotating this data. VARS supports the complete workflow from video ingestion and metadata management through taxonomic annotation to querying and data export. The system is freely available for use by other research institutions. 

The VARS stack is deployed via Docker using the varsq command-line orchestrator and consists of nine microservices handling annotation storage, video asset management, knowledge base taxonomy, image capture, API gateway, querying, and reverse proxy. Applications built on top of VARS include: 

  • VARS Annotation (GitHub) — Desktop application for creating and editing video annotations (Windows, macOS, Linux) 
  • VARS Query — Web application for searching and retrieving annotations, videos, and images 
  • VARS Knowledgebase Editor — Web application for editing the knowledge base (lexicon and phylogenetic tree of annotation terms) 
  • VARS Gridview — Bulk editing tool for reviewing and correcting bounding box annotations 
  • Mondrian — Image annotation application 

For more information about running MBARI’s VARS system: 

 

Midwater Time Series (901221) 

Between the surface of the sea and the ocean floor lies a vast fluid universe, Earth’s least-known environment. MBARI has sophisticated systems that have spent thousands of hours surveying and describing the deep waters of the ocean. In support of MBARI’s Midwater lab, Brian develops tools, technology, and analytical techniques for working with this large collection of data. 

 

FathomNet (901809) 

FathomNet is an open-source image database of expert-annotated underwater imagery used to train machine learning models for automated detection and classification of marine life. Brian contributes to the design and development of the FathomNet platform, including its core database backend and supporting microservices. 

The FathomNet Database provides a REST API for querying annotated images, taxa, and machine learning models. The API is documented via a Swagger/OpenAPI interface and is accessible programmatically via the fathomnet-py Python client library. 

Supporting infrastructure includes: 

  • worms-server — A high-performance in-memory server that exposes the World Register of Marine Species (WoRMS) taxonomic tree via a name-based REST API. Augments the standard WoRMS API with endpoints for full ancestor/descendant traversal, common-name lookup, and fast prefix/substring search. Also merges in MBARI Knowledgebase branches (equipment, geological features) for use in FathomNet annotation workflows. API docs at database.fathomnet.org:8888/docs/. 

 

Expedition Database Modernization (902403) 

The Expedition Database (EXPD) is MBARI’s core system for managing pre- and post-cruise metadata for all ship-based operations (mbari.org/cruises). Originally built in the late 1990s on Microsoft SQL Server with an Active Server Pages / Perl frontend, the system is now undergoing a full modernization to address 20+ years of accumulated technical debt and to support new requirements driven by the arrival of the R/V David Packard. 

The modernization effort spans three years (2024–2026) and covers two interconnected components: 

  • expd-rest-api — A Quarkus-based Java REST API that replaces the monolithic ASP/Perl web application. Built with Hibernate ORM/Panache and Jakarta REST, it exposes cruise and dive metadata programmatically — enabling downstream systems such as VARS to automatically retrieve expedition data without human mediation. The API is containerized and deployed via Docker. 
  • expd-support — A Scala 3 / sbt project containing database migration tooling. Uses ZIO for effect management and circe for JSON, and handles the schema migration from the legacy EXPD database to the redesigned schema with enforced foreign key constraints and support for new platform types (ships, ROVs, AUVs). 

Key accomplishments to date include a redesigned database schema, a deployed REST API serving expedition and dive records, and tested migration code for the legacy database. Planned work for 2026 includes a Vue.js web interface replacement, data load path migration (CTD, navigation), dive log app updates for shipboard compatibility, and full REST API integration test coverage. 

 

VARS-ML Vector Database (902604) 

This project extends MBARI’s Video Annotation and Reference System (VARS) by building a vector database to store image embeddings—mathematical “fingerprints”—for the hundreds of thousands of annotated deep-sea organisms in MBARI’s video archive. 

With this system, scientists can search for visually or semantically similar organisms instantly, cluster related observations, and even run natural language queries like “show me all the red fish.” This opens new avenues for species discovery, behavior analysis, and ML model development that aren’t possible with traditional database searches. 

The project runs 2026–2027, with year one focused on infrastructure and year two on expanding search and discovery capabilities for researchers across MBARI. 

Orenstein EC, Woodward B, Lundsten L, Barnard K, Schlining B and Katjia K (2025) Assisting human annotation of marine images with foundation models. Front. Mar. Sci. 12:1469396. doi: 10.3389/fmars.2025.1469396
 
Barnard, K., Liu, E., Walz, K., Schlining, B., Jacobsen Stout, N., & Lundsten, L. (2025). DeepSea MOT: A benchmark dataset for multi-object tracking on deep-sea video. arXiv:2509.03499. https://arxiv.org/abs/2509.03499
 

Borremans, Catherine, et al. “Report on the Marine Imaging Workshop 2022.” Research Ideas and Outcomes 10 (2024): e119782.

Katija, K., Orenstein, E., Schlining, B., Lundsten, L., Barnard, K., Sainz, G., Boulais, O., Woodward, B. and Bell, K.C., 2022. FathomNet: A global underwater image training set for enabling artificial intelligence in the ocean. Scientific Reports. https://doi.org/10.1038/s41598-022-19939-2

Katija, K., Schlining, B., Lundsten, L., Barnard, K., Sainz, G., Boulais, O., Woodward, B. and Bell, K.C., 2021. FathomNet: An Open, Underwater Image Repository for Automated Detection and Classification of Midwater and Benthic Objects. Marine Technology Society Journal, 55(3), pp.136-137. https://doi.org/10.4031/MTSJ.55.3.20.

Boulais, O., Woodward, B., Schlining, B., Lundsten, L., Barnard, K., Bell, K.C. and Katija, K., 2020. FathomNet: An underwater image training database for ocean exploration and discovery. arXiv preprint arXiv:2007.00114.

Schlining, K., Von Thun, S., Kuhnz, L., Schlining, B., Lundsten, L., Stout, N.J., Chaney, L. and Connor, J., 2013. Debris in the deep: Using a 22-year video annotation database to survey marine litter in Monterey Canyon, central California, USA. Deep Sea Research Part I: Oceanographic Research Papers, 79, pp.96-105. https://doi.org/10.1016/j.dsr.2013.05.006.

Paull, C.K., Schlining, B., Ussler, W.I.I.I., Lundste, E., Barry, J.P., Caress, D.W., Johnson, J.E. and McGann, M., 2010. Submarine mass transport within Monterey Canyon: Benthic disturbance controls on the distribution of chemosynthetic biological communities. In Submarine mass movements and their consequences (pp. 229-246). Springer, Dordrecht. https://doi.org/10.1007/978-90-481-3071-9_19.

Schlining, B.M. and Stout, N.J., 2006, September. MBARI’s video annotation and reference system. In OCEANS 2006 (pp. 1-5). IEEE. https://doi.org/10.1109/OCEANS.2006.306879.

Paull, C.K., Schlining, B., Ussler III, W., Paduan, J.B., Caress, D. and Greene, H.G., 2005. Distribution of chemosynthetic biological communities in Monterey Bay, California. Geology, 33(2), pp.85-88. https://doi.org/10.1130/G20927.1.

Graybeal, J., Gomes, K., McCann, M., Schlining, B., Schramm, R. and Wilkin, D., 2003, June. MBARI’s SSDS: operational, extensible data management for ocean observatories. In 2003 International Conference Physics and Control. Proceedings (Cat. No. 03EX708) (pp. 288-292). IEEE. https://doi.org/10.1109/SSC.2003.1224165.

Drazen, J.C., Goffredi, S.K., Schlining, B. and Stakes, D.S., 2003. Aggregations of egg-brooding deep-sea fish and cephalopods on the Gorda Escarpment: a reproductive hot spot. The Biological Bulletin, 205(1), pp.1-7. https://doi.org/10.2307/1543439.

Chavez, F.P., Pennington, J.T., Castro, C.G., Ryan, J.P., Michisaki, R.P., Schlining, B., Walz, P., Buck, K.R., McFadyen, A. and Collins, C.A., 2002. Biological and chemical consequences of the 1997–1998 El Niño in central California waters. Progress in Oceanography, 54(1-4), pp.205-232. https://doi.org/10.1016/S0079-6611(02)00050-2.

Chavez, F.P., Strutton, P.G. and Schlining, B.M., 2001. Bio-Optical Measurements at Ocean Boundaries in Support of SIMBIOS. SIMBIOS Project 2000 Annual Report, p.51.

Schlining, B., 1999. Seasonal intrusions of equatorial waters in Monterey Bay and their effects on mesopelagic animal distributions (Master’s thesis, California State University, Stanislaus).