SEA-SAAVI™, or waterside Simulation Assessment of Augmented/Autonomous Vessel Intelligence, is an evidence based simulation assessment program designed to evaluate the competency and capabilities of either assistive navigation technologies or autonomous navigation systems based on some level of engineered artificial intelligence (AI). Developed by members of the same team that created other global standard maritime skills assessments, SEA-SAAVI™ utilizes state of the art maritime simulation systems to provide a repeatable, objective measure of maritime skill performance compared to internationally established standards. SEA-SAAVI™ is founded on over 15 years of assessment development and use of the program to assess mariners throughout the globe, and to select the best possible candidates for marine pilots in major US ports. To date, other standardized assessment created by the team have sampled over 2000 mariners worldwide, and have provided insight to SEA-SAAVI™ about the typical performance of masters and mates the world over. The results of this past research work and adherence to standards forms the basis of recommendations and suggestions provided by the program. The target skill level is the managerial level, as established by the IMO and other flag state authorities, and aligns with the expectations and skills typically associated with a Master of an internationally sailing vessel.
The assessment itself is sourced and based on the full suite of capabilities required to safely navigate a vessel from berth to berth, and specifically focuses on International Regulations for Preventing Collisions at Sea 1972 (COLREGS), piloting and navigation, communication either internally or externally, use of tools and instrumentation, and the principles of good Bridge Resource Management (Human-Machine Interface) as it applies to any interface with potential crew members or local/remote supervisors. The standards of performance are derived from SCTW 2010 (International Convention on Standards of Training, Certification and Watchkeeping for Seafarers), COLREGS, Global Maritime Distress and Safety System (GMDSS), best practices sourced from a multitude of handbooks and guides, as well as internationally standardized procedures for traffic schemes, radio telephone communication, and tug/support vessel coordination. Outcomes of the assessment include a comprehensive report of performance compared to the standards describing specific category performance a Highly Effective, Effective, Not Effective or Unsatisfactory/Unsafe. The reports can also include performance recommendations, with suggestions for improving or modifying the outcome actions of the assistive technology.
Unique to this program is the ability to assess either a completely autonomous AI system (black box) or some form of intermediate combination of human mariner/supervisor and assistive technology hybrid, as in either case, the complete package constitutes the system under evaluation. And in the case of a hybrid solution, the program can evaluate the human mariner both separately and in combination with the technology in order to determine the relative level of skill improvement gained by introducing the technological solution.
We seek to offer this unique program to further develop relevant autonomous technologies presently being developed by leading edge technology and transportation companies. The insight gained through years of experience in this field, coupled with an objective, evidence based simulation, would provide a significant jump forward in understanding how well current autonomous technologies will function in the world of humans, and suggest a path forward for improvements or refinements to such systems. Ultimately, the success of automated or autonomous systems is the trust that humans put into the assistive technologies, and SEA-SAAVI™ hopes to provide a data based reason for placing trust in these new tools.