How do underwater robots achieve precise localization in deep-sea environments without GPS signals?

Underwater Robot
Views: 0Created At: 9/20/2025Updated At: 10/4/2025
楠 陈
楠 陈
PhD student researching advanced underwater manipulation.

Imagine you're in a pitch-black room, unable to see your hand in front of your face, and you want to know where you are. What would you do? You'd probably estimate your position based on how many steps you've taken and how many turns you've made, right? Underwater robots (ROVs or AUVs) in the deep sea are in a similar situation. GPS signals can't penetrate seawater, so they have to use some "black tech" to navigate.

This isn't achieved by a single technology; it's usually a combination of several methods working together like a team:

  1. Inertial Navigation System (INS) - The Robot's Own "Sense of Direction" This system is the core, like the robot's "cerebellum" and "inner ear," responsible for sensing its own attitude and motion. It contains gyroscopes (sensing rotation) and accelerometers (sensing acceleration/deceleration). After the robot starts from a known point, the INS continuously calculates how far it has moved in which direction and how many degrees it has turned. This is like you counting your steps and turns in a dark room.

    • Pros: Completely autonomous, requires no external signals, and provides real-time position.
    • Cons: Accumulates errors. Just like your estimation of distance and turns becomes less accurate the longer you walk, this error accumulates over time and distance, and the robot can slowly "get lost."

  2. Doppler Velocity Log (DVL) - The "Eyes" that Look at the Seabed To solve the INS error problem, the robot needs to "look down at the path" from time to time. That's what the DVL does. It emits several beams of sound waves towards the seabed and then receives the reflected signals. Based on the change in sound wave frequency (the Doppler effect, like the change in pitch of an ambulance siren as it passes you), it can very accurately measure the robot's speed relative to the seabed.

    • Function: The DVL tells the INS: "Hey, your last speed calculation was off; I have the accurate one!" This significantly reduces the accumulated error of the INS, allowing the robot to move more steadily and accurately between "calibrations."

  3. Acoustic Positioning System - The "Lighthouses" or "Satellites" of the Sea Relying solely on self-estimation and looking at the seabed isn't always enough; sometimes, external "coordinate references" are needed. This is where acoustic positioning systems come in, mainly in these types:

    • Long Baseline (LBL) System: This is the most accurate type. Several "acoustic transponders" (like "lighthouses" placed on the seabed that know their precise locations) are pre-deployed on the seabed in the operating area. The robot sends out an interrogation signal, and these "lighthouses" respond. By measuring the round-trip time of the signal, the robot can use triangulation to calculate its precise position relative to these "lighthouses." This is like being able to determine your position in a city by seeing three landmark buildings.
    • Ultra-Short Baseline (USBL) System: This system is more flexible. The acoustic transceiver is installed on the bottom of the mother ship, which has GPS on the surface, giving it an accurate position. The ship sends an acoustic signal to the underwater robot, and the robot responds upon receiving it. The receiver array on the ship can calculate the robot's distance and bearing relative to the mother ship based on the tiny time differences in receiving the response signal. The mother ship then "tells" the robot its calculated coordinates via acoustic communication. This is somewhat like the mother ship holding a GPS on the surface and extending an "invisible long pole" to constantly point at the underwater robot, telling it where it is.

  4. Terrain-Based Navigation (TBN) / SLAM Technology - The Art of "Recognizing the Way" This is a more intelligent method. As the robot moves, it uses sonar to scan the seabed topography, creating a "map." Then, it compares this real-time small map it has drawn with a pre-stored high-precision nautical chart of the area. By matching features like underwater mountains and valleys, the robot can "recognize" where it is. SLAM (Simultaneous Localization and Mapping) technology is even more advanced; it doesn't even need a pre-stored map. It can draw its own map while simultaneously determining its position on that map, much like an explorer charting an unknown territory as they explore.

In summary:

Precise underwater robot positioning is like a team effort:

  • Inertial Navigation (INS) is the main player, constantly estimating position but prone to errors.
  • Doppler Velocity Log (DVL) acts like a supervisor, continuously correcting the INS's speed to prevent it from going too far astray.
  • Acoustic Positioning Systems (LBL/USBL) are external "referees" that periodically provide an absolutely accurate position, bringing the INS back on track.
  • Terrain Matching (TBN/SLAM) is for advanced players, recognizing the way by observing the environment.

In practical applications, INS and DVL are usually tightly integrated (this combination is called an "inertial navigation unit"), and then periodically calibrated with an acoustic positioning system. This allows for long-duration, high-precision underwater positioning.

Created At: 09-20 17:51:28Updated At: 09-30 07:49:38