Method and System for Submarine Cable Path Planning

Link Copied.

Opportunity

Submarine fiber-optic communication cables are critical arteries of the global internet, carrying over 99% of international voice and data traffic. This vast network represents a massive financial investment, with costs reaching approximately $25,000 per kilometer. Despite their importance, these cables are vulnerable to various natural and man-made hazards, including earthquakes, volcanic eruptions, fishing activities, and anchor drops, leading to over 100 reported faults annually. Such damage or breaks can cause severe socio-economic consequences, especially for industries like finance where even brief communication outages result in significant losses. The current industry standard for designing reliable and cost-effective submarine cable routes relies heavily on manual methods and expert knowledge. This process involves using various data sources and tools to generate candidate paths, followed by feasibility checks and detailed comparisons. However, this approach is subjective, time-consuming, labor-intensive, and often fails to produce near-optimal paths due to a lack of scalability and precision. It cannot effectively balance cost and risk factors for ultra-long-distance cables exceeding 10,000 km, which are increasingly common. The challenge is exacerbated by the enormous scale of target areas and the massive datasets involved in high-precision planning, making traditional methods computationally infeasible or excessively slow.

Technology

This patent addresses the aforementioned problems by introducing an Adaptive Parallel Fast Marching Method (APFMM) for high-precision, ultra-long-distance submarine cable path planning. The core innovation lies in the novel integration of adaptive domain decomposition and multi-resolution analysis (MRA) with the Fast Marching Method (FMM). The FMM is a numerical technique for solving the Eikonal equation, which can find the minimum cost path between two points by considering a comprehensive lifecycle cost function that incorporates factors like cable length, seismic risk, seabed slope, and human activities. However, the sequential nature of the standard FMM poses computational bottlenecks for large-scale problems. The patented APFMM overcomes this by dynamically decomposing the target geographical area. It starts by decomposing the area into two initial subdomains along the direction connecting the start and end points and performs FMM in parallel to obtain an initial path. Crucially, it then adaptively refines the decomposition: it identifies a region containing the initial path, increases the data resolution within this region, and decomposes it into multiple subsequent subdomains aligned with the path direction. FMM is executed in parallel across all these subdomains. This process iterates, progressively increasing the number of subdomains and the resolution in critical areas near the evolving optimal path, while maintaining lower resolution in irrelevant areas. This adaptive strategy ensures load balancing among computational threads, minimizes communication overhead and "rollback" operations between subdomains, and focuses computational resources where they are most needed. The result is a highly scalable algorithm that can efficiently handle datasets for routes over 14,000 km, achieving automated, high-accuracy path planning that optimizes the total lifecycle cost.

Advantages

Solves the computational limitations of sequential FMM for large-scale, high-precision path planning over ultra-long distances (≥10,000 km).
Achieves significant speedup, with reported runtime reductions of over 81% compared to general parallel FMM methods.
Enables fully automated path planning, reducing reliance on subjective manual expert methods.
Improves path accuracy and optimality by systematically minimizing a comprehensive lifecycle cost function.
Enhances computational efficiency through adaptive domain decomposition that balances thread loads and minimizes inter-subdomain communication.
Dynamically focuses computational resources on critical path areas using multi-resolution analysis, avoiding unnecessary processing.
Provides a scalable and widely applicable solution for long-distance cable route planning.

Applications

Planning and optimization of submarine fiber-optic communication cable routes.
Path planning for other underwater infrastructure like power cables or pipelines.
Route planning for terrestrial infrastructure such as power transmission lines, gas pipelines, and road/rail networks.
Large-scale data processing and complex path planning in fields requiring optimization over extensive geographical areas.
Integration into commercial cable planning software tools to enhance automation and decision support.