Marine and underwater robotics

Autonomous Underwater Vehicles / AUVs

Untethered underwater robots for mapping, survey, and science missions.

Category reference

What autonomous underwater vehicles / auvs are

Autonomous Underwater Vehicles / AUVs are robotic systems built for offshore inspection, marine research, underwater mapping. They use sensors, actuators, embedded compute, control software, and task logic to act in physical environments.

This reference describes the category rather than a particular commercial product. Capabilities depend on the robot, its tools, software, operating environment and safety design. Product claims should be checked against the manufacturer documentation for the exact model and configuration.

How autonomous underwater vehicles / auvs work

A autonomous underwater vehicles / auvs system senses the world using sonar, DVL, pressure depth sensors, IMU, underwater cameras, estimates state, plans a task or route, and commands thrusters, buoyancy control, rudders, manipulator arms. Feedback loops compare the intended motion with what actually happened and trigger corrections, retries, or a safe stop.

Typical applications

  • offshore inspection
  • marine research
  • underwater mapping
  • aquaculture checks
  • ship hull inspection

Representative systems and formats

  • Autonomous Underwater Vehicles / AUVs research platforms
  • Autonomous Underwater Vehicles / AUVs commercial systems
  • Autonomous Underwater Vehicles / AUVs pilot deployments

Key technologies

  • pressure resistant design
  • acoustic navigation
  • thruster control
  • mission autonomy
  • waterproof electronics

Common sensors

  • sonar
  • DVL
  • pressure depth sensors
  • IMU
  • underwater cameras
  • acoustic positioning

Actuation and movement

  • thrusters
  • buoyancy control
  • rudders
  • manipulator arms
  • tether management systems

Software functions

  • acoustic navigation
  • mission planning
  • terrain following
  • sensor fusion
  • tether monitoring
  • underwater mapping

What to verify before deployment

A category description cannot predict performance in a specific workplace. Test the real task, environment and exception cases. Record where the system needs human recovery and confirm that the complete application has an appropriate safety assessment.

  • Performance drops when sensors face glare, dust, occlusion, deformable objects, poor lighting, water, smoke, or unexpected human behavior.
  • Hardware maintenance matters because motors, joints, seals, batteries, cables, and sensors degrade.
  • Most reliable autonomy is narrow and workflow specific.
  • Integration cost includes training, safety validation, spare parts, maps, network coverage, and support.
  • Human supervision is often needed for edge cases, recovery, cleaning, charging, or exceptions.

Continue with reviewed resources