TECHNOLOGY

The digital infrastructure powering software-defined ships
Every sensor, every signal, every system unified under one
AI-ready architecture, built and proven on the open ocean.
Wireframe blueprint diagram of the Liberty ship, labeled with Perception Suite, Electric Grid, and Diesel Engine Modules
Perception Suite
Automation Panels
Electric Grid
Diesel Engine Modules
Perception Suite Automation Panels Electric Grid Diesel Engine Modules
Re-Architecting Hull, Mechanical & Electrical (HM&E) from the Keel Up.
We redesigned ships from first principles, starting from the engine room out. Around our engine modules, we built a fully software-defined ship with smart, modular components designed to swap out for rapid repair. Together, they become building blocks for many ships, in many shipyards, for many missions.

The architecture meets and exceeds U.S. Navy requirements and is not limited to any single hull or program. Liberty Class is the first platform; the same foundation applies wherever autonomous ship operation is required.
See Liberty Class
Open Keel
The operating system for software-defined ships
Every sensor. Every signal. Every system. Unified.

Most ship automation is proprietary and siloed. Data gets locked in OEM-controlled black boxes. AI has nothing to reason over.

Open Keel is BWA's open architecture platform: the connective tissue integrating every subsystem on a vessel into a single, AI-ready data layer: propulsion, power generation, fuel management, sensors, bilges, navigation, machinery control. All of it flows through Open Keel, and every hour of operation makes the system smarter. Open Keel is the foundation every other capability in this stack is built on.
Engineering Autonomy
Autonomous Chief Engineer (A-CHENG)

A-CHENG is not a crewed-ship automation product added after the fact. It was designed from the start for a vessel with no one aboard: the decision-and-optimization engine that keeps making decisions when there is no crew to make them.

Automation monitors. Automation alerts. Automation stops. A-CHENG acts.
‍
The difference in practice:
01
Traditional automation
A fuel tank level sensor fails. The system loses its primary input, assumes the worst, halts operations, and alerts the crew  with no diagnosis and no lead. Hours lost chasing a false reading.
02
A-Cheng
Activates pumps to check pressure and flow, cross-references maintenance records, and reasons through a likely installation error all before anyone picks up the phone. Operations continue. The fix is targeted.
That kind of reasoning doesn't come from software alone. A-CHENG was built with mariners holding 100-ton licenses and decades of hands-on experience managing engineering plants at sea. Their knowledge of how ships actually fail, and how engineers actually diagnose them, is encoded into how A-CHENG thinks.
‍
Whereas automation follows a script, A-CHENG has an objective. It manages the whole ship's engineering plant, allowing it to run for months at a time at sea without human technicians, maintainers, or engineers.

Unlike legacy maritime automation systems that require yard visits and multi-year revision cycles to update, A-CHENG ships updates over the air and is built API-first anticipating operators who need a single view across a fleet, not one screen per vessel.


NERVE

A-CHENG is built on NERVE, Blue Water's purpose-built sensing and control layer. Where traditional marine automation relies on PLC and remote I/O systems from multiple vendors, NERVE unifies edge compute and control logic into a single platform across every sensor and actuator on the plant.

Unmanned operation requires denser sensor coverage, faster decisions, and direct control of valves and actuators, not just monitoring. NERVE is built for all three, with processing that happens at the edge and panels that update in the field, not in drydock.
Engine Modules
Containerized power. Swappable in minutes.
Blue Water ships are powered by diesel engine modules built into 20’ ISO containers. Multiple modules form a microgrid generating megawatts of power that drives everything from propulsion to payloads.
‍
Each module manages its own lube oil, filtration, and thermal systems autonomously. The grid architecture means the ship stays operational even if one module goes offline, enabling 2-3 months of no-touch operation at sea.
‍
When a vessel returns to port, the components most in need of maintenance are contained inside the module. Modules are replaceable in hours at any commercial port decoupling maintenance from the ship's operating schedule.
OPERATIONAL AVAILABILITY
90%
Designed to stay mission ready
Our system
90%
Conventional unmanned ships
65%
Crewed vessels
40%
System Validation
Build a Little. Test a Little. Learn a Lot.
Tow Tank
Hull designs are validated at MIT's tow tank alongside MIT's ocean engineers. For Liberty, this testing confirmed Sea State 7 survivability (6-9 meter waves) and 27-knot top speed, at scale, before steel is cut.
R&D Lab
Prior to hitting the water, we test critical components and systems in our lab on land. Cooling loops, electrical panels, and sensors all undergo hundreds of reps and sets in isolation, before integration onto the ATV or the ship.
Ironsides Autonomous Test Vessel (ATV)
100 feet and 170 tons, Ironsides validates BWA's full ship stack in real conditions: diesel-electric power generation, fuel management, cooling, bilge, propulsion, and navigation autonomy, all under continuous A-CHENG supervision. Every subsystem progresses through software-in-the-loop, hardware-in-the-loop, and open-water testing before ship integration.
Atlantic Endurance
82-foot crew transfer vessel upgraded with BWA's full navigation autonomy stack: 5 cameras, 2 radars, INS, full compute cluster, and navigation autonomy in partnership with MarineAI. The upgrade makes it the first offshore service vessel (Subchapter L) in the United States to receive U.S. Coast Guard approval for autonomous testing. On-water COLREGs testing takes place five days per week.
Built with the best