Incident Overview
On June 23, 2025, during testing at Channel Islands Harbor Marina in California, an unmanned surface vessel operated by the U.S. Navy received an unintended command. The vessel accelerated unexpectedly, striking and capsizing its crewed support boat. The incident resulted in the captain of the support boat being thrown into the water, though emergency responders recovered him without requiring medical attention. The Navy halted the trial immediately and coordinated with the Coast Guard and local harbor patrol for assistance.
The unmanned vessel was under tow when the unintended command was executed, leading to the collision. The Navy’s safety reporting system later disclosed the event, highlighting gaps in command validation within unmanned maritime systems.
The support boat, a tugboat, was part of the testing infrastructure, tasked with assisting in the deployment and recovery of the autonomous vessel. Its capsizing underscored vulnerabilities in the coordination between onboard control systems and external autonomy software, particularly in scenarios involving mixed human-machine operations.
No fatalities were reported, but the potential for injury or death remained a critical concern in the aftermath.
Possible Root Cause
Investigations into the incident identified two primary contributing factors: software failures and human error. The unintended command appeared to stem from a communication breakdown between the onboard vessel control systems and the external autonomy software managing the unmanned vessel’s operations.
The software’s inability to validate or reject anomalous commands highlighted a systemic flaw in the design of autonomous maritime systems.
Human error played a secondary but significant role in the incident.
The incident also revealed broader challenges in integrating unmanned systems with traditional maritime operations. The support boat’s role as a crewed vessel in close proximity to an autonomous platform introduced additional risks, particularly in scenarios where human and machine-controlled assets share the same operational space. The absence of standardized protocols for command validation and emergency intervention further compounded these risks.
Technical and Operational Context
The unmanned vessel involved in the incident was developed by BlackSea Technologies, a company known for its work in autonomous maritime systems. The vessel was equipped with onboard control systems designed to execute commands from external autonomy software, which in this case failed to prevent the unintended acceleration.
However, the incident demonstrated that such systems may not inherently account for human factors or environmental variables that could lead to unintended consequences.
During the trial, the unmanned vessel was under tow, a common practice in maritime testing to ensure stability and control before full autonomous operation. The capsizing of the support boat indicated that the autonomous vessel’s response to the command was both rapid and unchecked.
The Navy’s use of unmanned surface vessels in testing reflects a broader trend in military and defense technology toward increased automation. While such systems offer advantages in terms of reduced crew exposure and enhanced operational capabilities, they also introduce new risks, particularly when human and autonomous systems interact in dynamic environments.
Takeaways
The incident underscores the necessity for robust command-validation mechanisms in autonomous maritime systems. Current designs may prioritize functionality and efficiency over safety, particularly in scenarios where human operators and unmanned vessels share the same operational space.
Human error remains a persistent challenge in the integration of autonomous systems, even in highly controlled testing environments. Clear communication channels between human operators and machine systems are essential to prevent miscommunication and reduce the likelihood of similar incidents.
The involvement of multiple stakeholders, including autonomous vessel manufacturers, defense technology companies, and military operators, points to the need for collaborative safety standards. Navy, in conjunction with industry partners, should develop and enforce guidelines that address the unique risks posed by unmanned maritime systems.
The physical harm to the support boat’s captain, though not fatal, serves as a reminder of the potential consequences when autonomous systems fail. Property damage and the disruption of testing operations further illustrate the broader impact of such incidents on military readiness and resource allocation. Addressing these risks requires a proactive approach to safety, with a focus on both technological and procedural improvements.
For the autonomous maritime industry, this incident provides a case study in the challenges of scaling unmanned systems without adequate safeguards. Navy and other military organizations continue to explore the use of autonomous vessels, lessons from this event can inform the development of more resilient and secure systems.
Sources
Incident summary from the AI Incident Database, used under CC BY-SA 4.0.