The shipping industry has long been defined by a paradox: the vessels we trust most with global trade are still guided by systems where human error accounts for most accidents. This reality has accelerated interest in Maritime Autonomous Surface Ships, but as this review makes clear, autonomy is only as safe as the algorithms that plot the course. The paper offers a structured look at how route planning has evolved, dividing the field into baseline trajectory methods, environment-adaptive approaches, and multi-objective intelligent decision-making. For those of us tracking maritime safety, this classification is overdue. It gives researchers and operators a shared vocabulary to compare tools that have previously been discussed in scattered, inconsistent terms. The absence of such a framework has been a quiet barrier to progress, and this work addresses that gap directly.
What stands out is the acknowledgment that no single algorithm is sufficient. Baseline methods are reliable but rigid. Environment-adaptive techniques respond to changing conditions but often struggle with the complexity of real-world traffic. Multi-objective systems are promising for balancing fuel efficiency, safety, and regulatory compliance, yet they remain computationally demanding and difficult to validate at scale. The review's emphasis on integrating multiple algorithms rather than searching for a perfect standalone solution reflects a mature understanding of the problem. It is not about building a single omnipotent system; it is about engineering interoperability. This connects directly to earlier work we have covered on Evaluating Maritime Safety Actions: A Data-Driven Approach to Risk Reduction, where short-term safety interventions were assessed for their actual impact rather than their theoretical promise. Both studies share a common thread: the need to move beyond intuition and toward empirical, measurable outcomes.
The review also raises a critical point about the International Regulations for Preventing Collisions at Sea, noting that current encoding methods for these rules remain inconsistent across algorithms. This is not a technical footnote. COLREGs are the product of human judgment and interpretation, and translating them into machine-readable logic is fraught with ambiguity. The paper suggests that refining how these regulations are embedded into planning systems will benefit not only autonomous shipping but the entire global trade ecosystem. That is a practical insight worth emphasizing. For operators, this means that compliance cannot be treated as an afterthought or a checklist; it must be engineered into the decision-making process from the outset. Our related coverage on Optimizing Autonomous Ship Safety: Adaptive Collision Avoidance Through Bayesian Analysis reinforces this point, showing how timing and probabilistic reasoning are central to effective collision avoidance. The two pieces complement each other: one focuses on the mechanics of avoidance timing, the other on the broader classification and integration of planning methods.
The takeaway for our readers is straightforward: autonomous shipping will not advance through a single breakthrough but through the disciplined integration of validated, measurable approaches. The review's call for incorporating uncertainty into route planning is particularly important. Shipping routes are not static; they are shaped by weather, traffic, and unpredictable human behavior. Algorithms that assume a deterministic world will fail when they encounter one. The open question is not whether autonomous ships will arrive, but whether the research community will develop the shared standards needed to make them trustworthy. We would tell any reader asking about this paper that it is a necessary step toward that goal, not because it solves the problem, but because it finally gives us a map of where we are. The detail to watch is how quickly the industry moves from proposing integrated frameworks to implementing them in operational settings. That transition will determine whether this review remains a reference point or becomes a relic of early-stage thinking.