← Sticker Wall AGENT · DETAIL

Orca AI Maritime Waze: Fleet Co-Captain Network Generating $11.9M Annual Revenue

Workflow: SeaPod's five daytime cameras and three thermal imaging cameras fuse real-time video with radar, AIS, and GPS data on th

AGENT

Key Fields

FIELD STAMPS

IndustryFintech

RegionGlobal（以色列）

ScaleSME

ChannelOnline

🔧 Workflow

SeaPod's five daytime cameras and three thermal imaging cameras fuse real-time video with radar, AIS, and GPS data on the vessel's NVIDIA edge processor. The AI detects and classifies non-AIS small boats, buoys, whales, and other objects, calculates CPA/TCPA, prioritizes risks, and pushes collision avoidance recommendations to the bridge display while the captain retains final decision-making authority. Anonymous data is synchronously uploaded to the Co-Captain network, feeding back into personalized collision avoidance commands for every ship.

🛠 Setup Requirements

Requires collaboration between marine engineering and computer vision teams: installing SeaPod hardware on the compass deck, connecting radar, AIS, and GPS, deploying NVIDIA edge computing units for on-vessel real-time inference, and integrating with the FleetView shore-based platform for fleet monitoring. AI models must be trained for maritime scenarios such as nighttime and dense fog, then iteratively updated using real-ship data. Single-vessel integration and classification society approvals take weeks to months, large-scale fleet deployments take years, and individuals cannot independently complete the full hardware solution.

🧰 Toolchain

🔧 SeaPod Vision Suite (5 daytime cameras and 3 thermal imaging cameras)
🔧 NVIDIA Edge Processor (on-vessel real-time object detection and CPA/TCPA calculation)
🔧 Co-Captain Network (fleet-wide anonymous sharing of weather, congestion, and security threat data)
🔧 FleetView Shore Platform (real-time video streams, event alerts, and fleet trend analysis)

💰 Revenue

1. Commercial Shipowners (fleets such as Seaspan, EPS, Maran Tankers, Ionic) subscribe to SeaPod and FleetView per vessel (primary revenue): shipowners pay installation fees plus a per-vessel subscription fee. Exact pricing per vessel has not been publicly disclosed. With over 1,600 vessels deployed, 2024 annual recurring revenue is approximately $11.9 million, accounting for roughly 100% of annual revenue, translating to about $992,000 monthly (published by the company as of 2024); 2. Replicator Tier—Agency implementation and subscription revenue sharing for small and medium shipowners: individuals or small teams handle site selection and deployment for smaller shipowners, charging implementation service fees and taking a cut of subscriptions. Public figures for pricing and client counts are unavailable, and the total revenue contribution cannot be verified; 3. Insurer Installation Subsidy Channel Rebates: insurers sometimes subsidize installation costs, collected per subsidized vessel. Neither the number of subsidized vessels nor the rebate unit price has been made public, and their share of the total pie remains unclear; 4. Opportunity Item—External Licensing of the Co-Captain Data Network: sources indicate this network reduces near-miss incidents by 54%, saves $100,000 in fuel per vessel annually, and cuts carbon emissions by 195,000 tons. Anonymous navigation risk data can be licensed to insurers, ports, and classification societies. The scale of this licensing stream is based solely on media estimates, lacks independent verification, and its proportion cannot be traced.

💸 Cost

Major costs include camera and edge computing hardware procurement, vessel installation and retrofitting, and R&D team salaries, along with FleetView cloud computing power and data storage. Beyond subscription fees, sales cycles are long, upfront heavy-asset investment is high, and single-vessel hardware costs and installation fees are undisclosed.

⏱ Time Investment

The system provides 24/7 automated lookout with near-zero additional operational effort for captains and watchkeepers; operations require roughly several working days per week for event reviews, model feedback iterations, and fleet safety and fuel consumption reporting.

🚀 Getting Started

It is difficult for individuals to replicate the entire hardware solution. One can start by providing marine AI consulting or acting as an agent for small and medium shipowners to handle SeaPod and FleetView selection and deployment, earning implementation and subscription shares. Alternatively, use cameras and public AIS data to build and test a collision avoidance object detection demo, and establish classification society certification partnerships once proven. The first step is recommended to be studying Lloyd's Register's AI navigation sea trial evaluation standards and targeting a small or medium shipowner for a pilot.

🔑 Keys to Success

✅ Quantifiable ROI: convincing shipowner financial decision-makers with hard metrics such as saving approximately $100,000 in fuel per vessel annually and reducing near-miss incidents by over 37%
✅ Classification Society Certification Endorsement: Lloyd's Register sea trial verification showing a 98.6% recall rate and 94% precision serves as the ticket to enter commercial shipping lanes
✅ Co-Captain Network Effect: the more data fleets share, the more accurate collision avoidance predictions become, creating a first-mover moat through data compounding
✅ Non-AIS Target Detection Capability: ability to identify small boats, floating containers, buoys, whales, and other objects missed by radar and AIS, combined with 360° panoramic coverage to eliminate blind spots, forming a differentiated selling point against traditional navigation systems

⚠️ 风险

⚠️ Long classification society certification cycles and high per-vessel installation and retrofitting costs lead to slow expansion in the heavy-asset model. If insurance subsidies recede or major client concentration is too high, cash flow comes under pressure, alongside competition from systems in the same track like SMART-SEA.
⚠️ Commercial shipping prosperity and geopolitical freight rate fluctuations directly impact shipowner capital expenditures. During economic downturns, new vessel installations and subscription renewals may be delayed, causing new deployment speeds to fall short of expectations.
⚠️ False positives or missed detections by AI vision under extreme sea conditions or rain and fog occlusions may lead to liability disputes. Continuous real-ship data feedback is required to iterate models while retaining human lookouts as a safety net, and compliance and insurance liability boundaries are still evolving.

📌 Real Cases

📌 Seaspan Case: One of the world's largest container shipowners deployed SeaPod across 64+ vessels, reducing near-miss incidents by 37%, increasing minimum passing distance by 35%, saving approximately $100,000 in fuel per vessel annually, and reducing 500 tons of CO2; Maran Tankers reduced safety incidents by 74%.
📌 Eastern Pacific Shipping (EPS) Case: Expanded to dozens of vessels following a pilot, reducing near-miss incidents by 48%, increasing minimum passing distance by 10%, and achieving a crew satisfaction score of 9.0/10.
📌 Ionic Case: In the most complex navigation waters of the North Sea, near-miss incidents dropped by 22%, minimum passing distance increased by 20%, and fuel consumption decreased by about 3%, validating the dual value of collision avoidance and fuel saving in congested waterways.

Sources