Master's in Artificial Intelligence for Maritime Operations
This master's prepares you to lead the digital transformation of the maritime sector with artificial intelligence. You will learn to build marine data pipelines, develop computer vision and multisensor perception, forecast metocean conditions and optimize routes, predict machinery failures, optimize ports and terminals, give autonomy to marine robots, secure AI systems and run models in production with MLOps, combining theory with hands-on labs and real maritime use cases so that you can apply what you learn from day one.
Prior knowledge check
These are the skills this program gives you. Tick the ones you already master and we will tell you which part of the syllabus is for you.
You are starting from scratch in this subject. The program is designed precisely for that.
What is the Master's in Artificial Intelligence for Maritime Operations?
This master's covers the full life cycle of an AI solution at sea and in port, from the sensor to the decision. You will study marine data types such as AIS, VDR, ECDIS, radar, sonar, IMU, CTD, ADCP and machinery logs; protocols such as NMEA, Modbus, MQTT and OPC UA, time synchronization, real-time ETL and time series storage; supervised and unsupervised learning and operational, risk-oriented metrics; object detection and camera, radar and AIS fusion in fog, glare and darkness; metocean forecasting with statistical models, recurrent networks and temporal Transformers, uncertainty and go/no-go windows; anomaly detection and remaining useful life for propulsion and auxiliaries integrated with CMMS; ETA, berth and yard optimization; state estimation, control and mission planning for ROVs, AUVs and USVs with SIL and HIL simulation; and edge deployment, CI/CD for models and drift monitoring.
What sets this master's apart: it is not a general data science program nor one limited to autonomous navigation, but a broad, engineering-oriented program that applies AI across ships, ports, offshore operations and marine robotics; it treats safety, OT cybersecurity, adversarial threats, functional safety, model governance, ethics and project management as part of the job, so your solutions can be audited and trusted at sea; and it closes with a Final Project in which you define, build and validate an AI solution for a real maritime problem, with dataset, metrics and validation plan.
Program highlights
- Marine data engineering: AIS, radar, sonar, IMU, NMEA, MQTT and OPC UA pipelines with data quality and lineage.
- Computer vision and camera-radar-AIS fusion for detection and alerts in low-visibility conditions.
- Metocean forecasting, uncertainty, weather windows and route optimization for fuel and safety.
- Predictive maintenance: anomaly detection, remaining useful life and CMMS integration for marine machinery.
- Smart ports and marine robotics: ETA and berth optimization, ROV/AUV/USV autonomy and SIL/HIL simulation.
- MLOps, OT cybersecurity, functional safety, model governance and ethics, plus a Final Project on a real use case.
Syllabus of the Master's in Artificial Intelligence for Maritime Operations
11 modules. Select a module to see its content.
AI fundamentals, marine data and solution architecture for maritime operations
- 1.1. Overview of AI in the naval domain: use cases, technical limitations and operational feasibility criteria
- 1.2. Types of marine data: telemetry, oceanography, AIS, marine weather, sonar and imagery, and their impact on modeling
- 1.3. The data life cycle: capture, cleaning, labeling, balancing and quality control with traceability
- 1.4. AI architectures: edge on board, on-premises in port and hybrid cloud with latency requirements
- 1.5. Fundamentals of supervised and unsupervised learning and algorithm selection by mission
- 1.6. Performance metrics: precision, recall, F1, ROC-AUC and risk-oriented operational metrics
- 1.7. Experiment management: data and model versioning, reproducibility and change control
- 1.8. Modeling risks at sea: shifting distributions, sensor noise and biases caused by environmental conditions
- 1.9. Technical documentation: data sheets, model cards and reporting of assumptions for audit
- 1.10. Integrative workshop: defining a naval use case with requirements, minimum data and acceptance KPIs
Marine data engineering, sensor integration and automated collection pipelines
- 2.1. Integrating sources: AIS, VDR, ECDIS, radar, IMU, CTD, ADCP and machinery logs
- 2.2. Protocols and formats: NMEA, Modbus, MQTT, OPC UA, APIs and schema normalization
- 2.3. Time synchronization: NTP and PTP, time-stamping, drift and multisensor consistency
- 2.4. Real-time ETL and ELT: parsing, enrichment, validation and error handling
- 2.5. Time series storage: modeling, partitioning, retention and compression
- 2.6. Data lakes and lakehouses for maritime operations: data governance, lineage and access control
- 2.7. Field data quality: detection of outliers and gaps, calibration and operational QA/QC
- 2.8. Pipeline observability: logs, metrics, traces, latency and SLA alerts
- 2.9. Data security: encryption, segmentation, hardening and service continuity at sea
- 2.10. Lab: designing an end-to-end pipeline from sensor to dashboard with validation rules
Computer vision and multisensor perception for navigation, safety and maritime surveillance
- 3.1. Fundamentals of computer vision: detection, segmentation, tracking and motion estimation
- 3.2. Detection of maritime objects: vessels, buoys, obstacles and wildlife in low visibility
- 3.3. Camera-radar-AIS fusion: target association, ambiguity resolution and state consistency
- 3.4. Image processing in extreme conditions: haze, glare, night, rain and vibration on board
- 3.5. CNN and Transformer models for perception: trade-offs between accuracy, latency and power consumption
- 3.6. Supervision and labeling: semi-automatic strategies, active learning and bias control
- 3.7. Evaluating perception: mAP, IoU, robustness by scenario and operational safety metrics
- 3.8. Edge deployment: quantization, pruning, accelerators and performance profiles on embedded hardware
- 3.9. Security and privacy: image handling, retention, anonymization and access policies
- 3.10. Applied project: a detection and alert system with multisensor integration and technical report
AI for oceanographic and marine weather forecasting and operational decision-making
- 4.1. Key metocean variables: wind, waves, currents and tides and their impact on operations
- 4.2. Time series: forecasting, seasonality, regime changes and window-based validation
- 4.3. Forecasting models: ARIMA and Prophet, recurrent networks, temporal Transformers and hybrid approaches
- 4.4. Data assimilation and fusion: buoys, satellites, numerical models and onboard sensors
- 4.5. Operational risk prediction: go/no-go thresholds, weather windows and safety margins
- 4.6. Route optimization: cost, time, comfort, fuel consumption and avoidance of adverse conditions
- 4.7. Uncertainty and confidence: intervals, probabilistic calibration and decisions under uncertainty
- 4.8. Smart alarms: detection of oceanographic anomalies and operational triggers
- 4.9. Integration with SOPs: procedures, escalation, evidence and decision logs
- 4.10. Case study: mission planning with metocean forecasting and acceptance criteria
AI for marine machinery, predictive maintenance and critical asset management
- 5.1. Propulsion and auxiliary systems: instrumented variables, failure modes and criticality
- 5.2. Condition modeling: vibration, temperature, pressure, consumption and degradation signatures
- 5.3. Anomaly detection: thresholds, statistical models, autoencoders and an asset-based approach
- 5.4. Failure prediction: remaining useful life (RUL), survival analysis and risk-based maintenance
- 5.5. Integration with CMMS and EAM: work orders, spare parts, SLAs and closure with evidence
- 5.6. AI-assisted root cause analysis: correlations, events, sequences and prioritization of actions
- 5.7. Optimizing shutdowns: windows, resources, cost and minimal operational impact
- 5.8. Validation ashore and at sea: testing, acceptance, false positives and iterative improvement
- 5.9. Applied OT cybersecurity: segmentation, least privilege and immutable event logs
- 5.10. Workshop: designing a predictive maintenance program with KPIs and an implementation plan
AI for maritime logistics, terminals, smart ports and the supply chain
- 6.1. Port and terminal processes: arrival, berth planning, yard, gate and hinterland
- 6.2. ETA/ETD and congestion prediction: models, features, drift and seasonal evaluation
- 6.3. Berth and resource optimization: allocation, sequencing and scenario simulation
- 6.4. AI for yards and equipment: moves, availability, maintenance and fewer rehandles
- 6.5. Performance analytics: operational KPIs, bottlenecks and Lean-style improvements
- 6.6. Pricing and revenue management: demand, elasticity, SLAs and penalties with a data-driven approach
- 6.7. Documentation and traceability: events, APIs and EDI, data quality and audit
- 6.8. Conceptual digital twin: a twin of the port or terminal to test policies and resilience
- 6.9. Risk and continuity: contingencies, systemic failures and operational recovery plans
- 6.10. Case study: a decision dashboard with prediction and resource optimization for a terminal
Intelligent marine robotics: ROVs and AUVs, autonomy, mission planning and control
- 7.1. Types of marine robots: ROVs, AUVs, USVs and mission-oriented hybrid architectures
- 7.2. State estimation: IMU-DVL-pressure fusion, relative navigation and graceful degradation
- 7.3. Motion control: advanced PID, thrust allocation, saturation and stability in 6 DOF
- 7.4. Mission planning: waypoints, geofences, constraints and abort and safe-return criteria
- 7.5. Underwater perception: sonar, vision, backscatter and basic detection and classification
- 7.6. Communications and telemetry: tether, acoustics, latency, link loss and store-and-forward strategies
- 7.7. SIL and HIL simulation: scenarios, fault injection and performance metrics
- 7.8. Operational safety: risk matrices, procedures, permits and coordination with the support vessel
- 7.9. Mission data management: logs, synchronization, evidence and inspection reports
- 7.10. Applied project: design and validation of a robotic mission with a complete technical deliverable
Cybersecurity, functional safety and resilience of AI systems in the naval environment
- 8.1. Attack surface of maritime systems: IT and OT, sensors, links and control stations
- 8.2. Threats to models: data poisoning, adversarial attacks, information leakage and input manipulation
- 8.3. Security by design: hardening, access control, segmentation, key rotation and logging
- 8.4. Resilience and continuity: redundancy, failover, backups, recovery testing and safe degradation
- 8.5. Functional safety: limits, interlocks, watchdogs and safe stop modes in autonomous systems
- 8.6. Incident management: detection, containment, forensic analysis and lessons learned
- 8.7. Model governance: approvals, changes, audit, traceability and deployment criteria
- 8.8. Technical and operational compliance: evidence, records, retention and supplier control
- 8.9. Risk assessment: matrices, thresholds, prioritization and acceptance by criticality
- 8.10. Workshop: a cybersecurity and resilience plan for an onboard or port AI system
MLOps, model operations, observability and continuous improvement in maritime production
- 9.1. End-to-end MLOps: development, validation, deployment and operation with version control
- 9.2. CI/CD for AI: testing, packaging, data validation and quality gates
- 9.3. Model monitoring: drift, performance decay, alerts and governed retraining
- 9.4. Pipeline observability: latency, availability, errors, backlog and SLA compliance
- 9.5. Experiment management: tracking, comparisons, reproducibility and documentation
- 9.6. Edge MLOps: deployment on board, secure updates and remote management
- 9.7. A/B testing and operational evaluation: real impact, safety, costs and operator experience
- 9.8. Dataset management and continuous labeling: active learning, data augmentation and bias control
- 9.9. Executive reporting: KPIs, risks, ROI and prioritization of the improvement backlog
- 9.10. Lab: implementing an MLOps cycle with monitoring and controlled retraining
Strategy, regulation, ethics and project management for AI in the maritime industry
- 10.1. Maritime AI roadmap: prioritization by value, risk, complexity and data maturity
- 10.2. Project management: scope, schedule, budget, risks and acceptance criteria
- 10.3. Procurement and suppliers: specifications, SLAs, ownership of data and models and validation of deliverables
- 10.4. Ethics and responsibility: bias, explainability, transparency and the limits of automation
- 10.5. Data governance: privacy, retention, access, transfers and internal policies
- 10.6. Change management: operational adoption, training, roles and organizational resistance
- 10.7. Designing AI-assisted SOPs: integration with operations, escalation and traceability
- 10.8. Measuring impact: ROI, TCO, residual risk, safety and business continuity
- 10.9. Portfolio of use cases: surveillance, navigation, maintenance, logistics and exploration
- 10.10. Final Project preparation: problem definition, methodology, dataset, metrics and validation plan
Final Project
- 11.1. Choosing the topic and defining the problem
- 11.2. Setting objectives, hypotheses and scope
- 11.3. Review of the state of the art and specialized literature
- 11.4. Analysis of the applicable regulatory and technical framework
- 11.5. Designing the methodology and work plan
- 11.6. Collection, processing and validation of data
- 11.7. Technical development and analysis of results
- 11.8. Discussion, conclusions and future lines of work
- 11.9. Writing the final report
- 11.10. Presentation and defense before an evaluation panel
Duration and format of the Master's in Artificial Intelligence for Maritime Operations
It is 1,500 hours. Tell us how many you can study each week and we will tell you when you will really finish.
This is the pace at which most students finish. It fits well with the installment plan.
Academic leadership and faculty
The Navalis teaching team: working professionals who teach and supervise the programs.
Dr. Marina AlcoverAcademic Director and Head of Quality Assurance
Eng. Diego CifuentesSenior Lecturer in Naval Welding and Non-Destructive Testing; Head of QA/QC for Welded Joints
Capt. Javier Abaroa (MCA)Senior Lecturer in Ship Handling, Electronic Navigation (ECDIS/Radar/ARPA) and GMDSS; STCW Compliance and Bridge Resource Management Coordinator
Dr. Pau FerrerSenior Lecturer in Naval Architecture and Simulation; Director of the Hydrodynamics and CFD Laboratory
Dr. Nuria LlobregatSenior Lecturer in Marine Meteorology, SAR and Operational Safety; Navigation Risk Management Coordinator
Eng. Javier BañulsSenior Lecturer in Electrical Engineering, Automation and ETO Engineering; Electronic Integration and Engine Alarms Coordinator
Eng. Sofía MarquinaSenior Lecturer in Materials, Composites and Corrosion Protection; Coordinator of the Advanced Paints and Coatings Line
Eng. Tomás RieraSenior Lecturer in Propulsion, Alternative Energy and Hybrid Systems; Energy Performance and Vibration CoordinatorObjectives of the Master's in Artificial Intelligence for Maritime Operations
Career opportunities after the Master's in Artificial Intelligence for Maritime Operations
Questions about the Master's in Artificial Intelligence for Maritime Operations
The maritime sector as a whole: ships and their machinery, navigation and surveillance, ports and terminals, offshore operations, oceanographic forecasting and marine robotics.
Autonomous navigation is only one of its applications. This program covers the full range of maritime AI, from data engineering and computer vision to predictive maintenance, smart ports, metocean forecasting, marine robotics, MLOps and the governance of AI systems.
Yes. You need programming fundamentals, ideally in Python, plus basic linear algebra, statistics and probability. Prior machine learning experience helps but is not required.
Yes. The labs and projects use maritime data types such as AIS, radar, sensor and machinery data and common tools for machine learning, data pipelines and model deployment.
Yes. A full module covers the attack surface of maritime IT and OT systems, adversarial threats to models, functional safety, resilience, incident management and model governance.
Yes, if you have or are willing to build the required programming and math foundations. Officers bring valuable operational knowledge, and the admission interview confirms that the program fits your profile.
Through labs, applied projects, workshops and short tests in each module, and a Final Project in which you build and validate an AI solution for a maritime problem.
A Navalis Institute master's certificate and a portfolio of pipelines, models, reports and your Final Project that you can show to employers.
More programs in Innovation and startups
Admission process
No waiting for a call back. If your documents are in order, you can enroll in the same session.
- Profile: degree in computer science, engineering, mathematics, physics or a related field, or demonstrable experience in software development or data analysis; programming fundamentals (preferably Python), linear algebra, statistics and probability are needed, and basic machine learning knowledge is an advantage.
- Language: sufficient command of English to follow the classes and the technical documentation.
- Technical resources: computer with a high-speed internet connection and the ability to install development tools such as Python, TensorFlow or PyTorch and work with virtual environments.
- Admission: online application, review of your profile and a short interview with the academic team.
- 01ApplicationYour details, background and documents.
- 02Instant reviewWe check that everything is complete and in order.
- 03EnrollmentChoose full payment or instalments and reserve your place.
Admission is provisional until the committee validates the documents provided. If you do not meet the requirements, the full amount is refunded.
Ask us before you enroll
Leave your details and admissions will write to you to answer any questions. If you are already sure, you can apply and pay in the same session.