Mythos AI’s APAS Enhances Maritime Radar Navigation

What Is APAS?

APAS (Advanced Pilot Assistance System) is a next-generation navigation support system developed by Mythos AI, an American maritime technology company. Its purpose is to assist mariners — especially on bridges of commercial vessels — by integrating data from radar, machine vision, ship dynamics and other sensors to provide prioritized alerts and actionable decision-support, rather than automating or replacing the human crew. 

Key design principles:

• Radar-first perception: Unlike many navigation assistance systems that primarily rely on cameras or other vision systems, APAS gives primary importance to radar data. Radar is robust in low visibility (fog, rain), at night, or when camera vision is limited.
  
• Multi-sensor fusion: It also uses machine vision and other sensors, but these are built around the radar backbone.
  
• Alert prioritization: The system processes multiple streams of data (radar, vision, ship performance) and delivers a single, prioritized alert to reduce distractions. The intent is to reduce cognitive burden on the bridge team.
  

Deployment Aboard CB Pacific

The CB Pacific is a chemical tanker operated by CB Tankers, which is part of the Lomar group. This vessel has been selected as a test platform for APAS. 

• The installation aboard CB Pacific was completed in early September 2025.
  
• The trial is year-long, intended to test APAS under real-world operational, environmental, and navigational conditions.
  
• CB Pacific was chosen because of its relatively predictable routes, stable radar setup (Furuno radar), and possibly because chemical tankers often have well-defined navigation constraints, which helps in establishing consistent test metrics.
  

How APAS Works: Radar-First Perception + Machine Vision

To understand what makes APAS different, we need to unpack its architecture and how it perceives, processes, alerts, and supports human decisions.

Radar-First Perception

• Radar sensors are robust to environmental conditions that degrade visibility: rain, fog, spray, low light. They detect range, bearing, relative velocity of objects (other vessels, obstacles).
  
• Mythos AI integrates directly with existing ship radars (e.g. Furuno models), meaning APAS doesn’t require replacement of primary radar hardware in many cases, easing deployment.
  

Sensor Fusion & Machine Vision

• Radar tracks surrounding objects, while machine vision classifies them—identifying vessel types, interpreting environmental context like lighting and waterline, and more.
• The fusion helps mitigate weaknesses of vision (low light, glare, occlusion) and of radar (less fine detail).
  

Intelligent Alerting & Bridge Integration

• APAS is not autonomous navigation; it does not steer or issue commands automatically under the current deployment model. Instead it generates prioritized alerts for bridge crew.
  
• The system takes into account vessel dynamics (speed, maneuverability, current trajectory) and environmental factors to assess collision risk or dangerous scenarios.
  

Trial Goals & Regulatory Compliance (COLREG)

A central part of the APAS trial aboard CB Pacific is ensuring regulatory compliance and real-world safety.

• COLREG: The Convention on the International Regulations for Preventing Collisions at Sea is the standard reference. The trial will evaluate APAS’s ability to align with COLREG rules and ensure that its alerts and suggested responses are consistent with international navigation law.
  
• Real-world conditions: The trial will test signal processing, collision prediction, and the effectiveness of alerting under varying conditions (weather, visibility, traffic) in live sea journeys.
  

Benefits: Safety, Situational Awareness, Crew Assistance

Based on the deployed features and objectives, here are expected benefits—some immediate, some over a longer horizon.

Enhanced Safety & Reduced Risk

• Early detection of collision risk gives the crew more margin to act.
  
• Better monitoring of proximity to other vessels, obstacles, navigation markers.
  

Reduced Cognitive Load

• Multiple data streams tend to overwhelm human monitoring; consolidating alerts into prioritized signals reduces distractions.
  

More Reliable Performance in Adverse Conditions

• Radar’s robustness means the system will function better in low visibility.
  

Preservation of Human Judgment

• By design, APAS supports rather than replaces the crew. This retains human decision-making, which is essential for legal, ethical, and practical reasons in maritime navigation.
  

Long-Term Operational Insights

• Over the year-long trial, data collected will include master mariner behavior, environmental norms, port approaches, etc. These datasets can improve model performance and calibrations for real vessels.
  

Challenges, Risks, and Human-in-the-Loop Considerations

Technology like APAS brings promise—but also risks. A realistic assessment must include what could go wrong, and what needs to be managed.

False Positives / False Negatives in Alerts

• If alerts are too frequent or too many false alarms, crew may suffer “alert fatigue” and begin to ignore or distrust the system.
  
• Conversely, missing a collision risk or hazard is dangerous and potentially catastrophic.
  

Sensor Limitations & Integration Issues

• Radar hardware may have limitations in resolution or update rate; radar clutter (sea waves, weather) can produce noise.
  
• Engineers must synchronize and calibrate integration with other sensors, as misalignments and time lags between radar and vision systems can produce errors.

Legal & Liability Concerns

• If APAS issues an alert and a crew neglects it, or misinterprets it, where lies liability?
  
• In case of accidents, regulators, insurers will examine how well APAS was calibrated, supervised, and how crew training was done.
  
• Compliance with COLREG is necessary but may need additional standards from classification societies (e.g., IMO guidelines, regional maritime authorities).
  

Human Trust, Training, and Culture

• Crew must trust the system; that requires good training, clear UI/UX, proven reliability.
  
• Culture in many maritime operations is conservative; adoption may be slow unless benefits clearly outweigh risks.
  

Environmental & Operational Variability

• Sea routes vary—weather, traffic, port stance, currents. The system must adapt to diverse scenarios, which may reduce consistency.
  

Operational Testing: Southern Devall & Inland Waters

Before the CB Pacific deployment, Mythos had tested APAS in inland waters:

• The Southern Devall towboat on the Mississippi River was the site of an earlier APAS deployment (around August 2025). This tested APAS in river navigation, including hazards like floating debris, displaced buoys, variable water depths and frequent maneuvers.
  
• Data from that helps refine APAS alert thresholds, behavior models, and understanding of how humans interact with assistance systems under different navigation regimes.
  
• These inland trials are critical, because they allow testing in complex but controlled environments; lessons learned there feed into open-sea deployments like CB Pacific.
  

Strategic Implications for the Maritime Industry

The APAS aboard CB Pacific trial has broader implications:

Commercial Shipping Operators

• Potential to reduce accidents, reduce insurance costs, reduce human errors.
  
• May enable more efficient navigation (optimizing speed, routes) and better fuel usage, indirectly yielding environmental benefits.
  
• Enhanced decision-support could extend useful life of vessels and reduce maintenance or wear related to navigational risk.
  

Startups & Maritime Tech Ecosystem

• Mythos AI, with its radar-first architecture, demonstrates a different design path from those relying heavily on cameras or purely vision systems.
  
• Partnership with established fleet operators (Lomar / CB Tankers) and innovation labs (lomarlabs) shows importance of operational access and real vessel data in validation.
  

Crew & Training

• Capturing mariners’ expertise and navigational norms will help in training younger or less experienced crew.
  
• Systems like APAS could become part of the standard bridge toolkit, requiring crew education, adaptation, and trust building.
  

Defence, Fleet-Scale, and Regulatory Landscape

APAS could appeal beyond commercial shipping:

• There is defence interest because precise, reliable navigation assistance is critical in strategic fleet operations. Radar-based, robust systems are especially relevant in low-visibility operations, contested environments, or where concealment or operational secrecy may limit reliance on external infrastructure.
  
• For fleet-scale adoption, adoption by classification societies (like Lloyd’s Register, DNV, ABS), regulatory approvals (IMO, local port authorities), and standards compliance will be key.
  
• Legal frameworks will need to define how assistance systems fit into bridge team operations, responsibilities, and liability.
  

Future Outlook & What Success Will Look Like

What would “success” for APAS look like a year or two from now? Key markers include:

• Demonstrated reduction in near-misses or navigational incidents aboard CB Pacific during trial, especially under adverse conditions.
  
• Strong crew feedback: usability, clarity of alerts, trust, minimal false alarms.
  
• COLREG compliance without ambiguity in alerting or suggestions.
  
• Regulatory acceptance and possibly certification by class societies or national maritime authorities.
  
• Expansion of deployments into other ship types, routes, and enviro-conditions (coastal, open sea, ports).
  
• Economic benefit: whether improved safety and efficiency translate into cost savings (fuel, insurance, downtime).
  

Conclusion

Mythos AI’s Advanced Pilot Assistance System (APAS) installed aboard CB Pacific marks a substantial step forward in maritime navigation assistance. It represents a maturing of the idea that crew support systems — not just automation — can significantly augment safety and performance at sea.

By choosing a radar-first perception system, integrating with existing radar hardware, fostering collaborations (such as with lomarlabs, CB Tankers), and deploying in real-world settings (inland rivers, tankers with known routes), Mythos AI is building a foundation for what could become standard bridge intelligence.

Nevertheless, the journey ahead involves addressing challenges: ensuring alerts are reliable and trusted, handling legal liability, training crews, and navigating regulatory frameworks. If the industry successfully overcomes these challenges, it may usher in a new era where human mariners and AI systems collaborate in close, trusted partnership — enhancing operational safety, reducing risk, and sharpening the intelligence of bridge technology.