Underwater Robot Cable Assembly: A Subsea Wiring Guide

In June 2026, a University of Florida team unveiled BlueME, a compact magnetoelectric-antenna system that let underwater robots talk to each other over 730 meters of seawater on about 10 watts. It is a genuine breakthrough for the hardest problem in marine robotics: communication. But read the fine print and the lesson for hardware engineers is the opposite of wireless hype. BlueME tops out in the kilobits-per-second range. The moment an underwater robot needs real power or live video, it is back on a cable. Underwater robotics is, first and last, a cable-assembly problem.

This guide is for engineers and sourcing teams building remotely operated vehicles (ROVs), autonomous underwater vehicles (AUVs), or subsea robotic tooling who have to get power, signal, and data through seawater reliably. We cover the umbilical/tether, sealing that actually survives depth, wet-mate vs dry-mate connectors, corrosion, and pressure. The wireless link is the headline; the wiring is the lifeline.

In a 2025-2026 program from our case bank, a customer in a harsh-environment sector pushed a custom harness into specialized braided cable with exact core colors, AWG, and stripe patterns, plus laser-etched marking and injection-molded connectors. The takeaway transfers straight to subsea work: in environments that punish wiring, the molded, sealed, fully-specified assembly is the product, and the supplier who can prove that capability earns the order.

TL;DR

• Underwater robots still depend on cable: wireless subsea comms (e.g. BlueME) is low-bandwidth, so power and live video ride a tether.
• An ROV umbilical bundles power, data (often fiber), and a strength member with managed buoyancy — it is a system, not a wire.
• IP68 alone is not a depth rating; hydrostatic pressure demands overmolding, potting, and water-blocking, validated by pressure test.
• Choose wet-mate connectors for in-water mating, dry-mate for sealed-on-deck assemblies — they are not interchangeable.
• Seawater drives corrosion and galvanic attack; material and shield/ground choices decide service life.
• Validate with IP, hipot, and immersion/pressure testing, not just continuity.

Why underwater robots are a cable problem first

Underwater robots split into two wiring worlds: tethered ROVs and free-swimming AUVs. An ROV hangs on an umbilical that delivers power and high-bandwidth data from the surface. An AUV carries its own battery and runs internal harnesses, surfacing or using short-range links to report. Both live or die by sealed interconnect.

The communication breakthroughs do not remove the cable; they reframe it. Acoustic links carry far but slowly and disturb marine life; optical links are fast but blind in turbid water; magnetoelectric systems like BlueME are compact and robust but measured in kilobits. None of them move kilowatts of thruster power or a 1080p inspection feed. That is why the tether persists: it is the only path that carries serious power and bandwidth at once.

Customers come in asking how to go wireless underwater, and the honest answer is: for telemetry, maybe; for power and live video, you are keeping the tether. The engineering that pays off is making that tether and its connectors survive the sea, not wishing it away.

— Hommer Zhao, General Manager and Wire Harness Engineer

The umbilical/tether: an ROV's lifeline

An ROV umbilical is a multi-element assembly, not a single cable. It bundles power conductors, data lines (frequently optical fiber for video and high-speed sensor returns), sometimes hydraulic or pneumatic lines, and a load-bearing strength member, all under a jacket engineered for a target buoyancy in water. Get the buoyancy wrong and the tether either drags the vehicle down or floats into the work site.

The build problems are termination and strain. Every conductor and fiber has to break out and terminate into a sealed bulkhead without letting water track along the cable. The strength member must take the tow load so the conductors never do. This is why the umbilical is specified as a system with the connector and the vehicle penetration, not bought as a length of wire. Our custom cable assemblies and power distribution harness work covers the power-plus-data breakout side of this.

Sealing: IP68 is not a depth rating

The most common subsea wiring mistake is treating an IP68 rating as proof a connection survives depth. The IP code IP68 means protection against continuous immersion under conditions the manufacturer specifies, often just 1.5 meters. A robot working at 100 meters sees roughly 11 bar of hydrostatic pressure forcing water into every seam. Depth rating is a separate, tested spec.

Surviving depth means designing the seal, not just selecting a connector. Overmolding (a fully molded, homogeneous boot over the termination), potting the backshell, and water-blocking the cable core so a nicked jacket cannot flood the whole run are the real defenses. For the rating framework see our robot cable assembly IP ratings guide; for build, our waterproof robot cable assembly and molded cable assemblies are purpose-built for this.

Wet-mate vs dry-mate connectors

Subsea connectors come in two families that are not interchangeable. Dry-mate connectors are sealed and mated in air (on deck or in a sealed housing) and must never be connected underwater. Wet-mate connectors are engineered to be mated and de-mated submerged, excluding water from the contacts as they join — essential for ROV tool changes, subsea docking, and reconfiguring equipment on the seabed.

The choice drives cost and design. Wet-mate is far more expensive and complex; use it only where in-water mating is genuinely required. Where the assembly is sealed topside and stays sealed, a robust dry-mate or fully molded penetrator is lighter and cheaper. For connector selection and sealed terminations, see our custom connector solutions.

Corrosion and materials in seawater

Seawater is a conductive, corrosive electrolyte, so material choice decides service life. Galvanic corrosion attacks the junction of dissimilar metals — a stainless backshell against a different alloy contact will corrode at the interface unless the pairing and any sacrificial protection are planned. Jackets must resist salt, UV at the surface, and abrasion against hull and seabed.

• Use jackets rated for seawater and abrasion (PUR/TPU families are common) rather than generic PVC.
• Plan dissimilar-metal junctions deliberately; isolate or match alloys to limit galvanic attack.
• Specify shielding and a defined ground/return so the conductive medium does not corrupt signal.
• Treat every exposed metal part (connector shells, clamps) as a corrosion candidate, not just the cable.

Pressure and depth engineering

Beyond sealing, depth changes how the assembly is built. Deep systems often use pressure-balanced, oil-filled designs so internal pressure tracks the outside and no air gap can be crushed. Water-blocking compounds inside the cable stop a single breach from flooding the length. The deeper the duty, the more the cable, connector, and vehicle penetration must be engineered as one pressure system.

This is also where reliability economics bite: a failed subsea connector can mean a vehicle recovery costing far more than the part. That asymmetry is why subsea buyers pay for validated, over-built assemblies and thorough wire harness testing rather than the cheapest connector.

Internal wiring of AUVs

A free-swimming AUV is, electrically, much like any mobile robot — battery, thruster motor power, encoder and sensor feedback, compute — except every run lives inside a sealed pressure hull, and any hull penetration (for a sensor, antenna, or charge port) is a flooding risk. The internal harness discipline mirrors land robots; the penetrations are the subsea-specific hazard.

So AUV wiring borrows directly from the rest of robotics: clean power distribution from the battery, shielded sensor and signal cables, and disciplined routing — then adds sealed, depth-rated penetrators wherever a conductor crosses the hull. For the broader robot wiring foundations, the robot arm internal harness approach applies inside the hull.

Testing and qualification

Continuity is the floor, not the proof, for subsea assemblies. A connection can pass continuity and still flood at depth. Require, per assembly: insulation resistance and hipot, an IP/immersion test, and a hydrostatic pressure test to the rated depth, ideally with the assembly powered and monitored during pressurization. Add corrosion/salt-spray exposure where service life matters.

Source a partner who can run and document these, not just crimp. Our wire harness testing and robot cable supplier qualification workflows cover the evidence a subsea program needs before it trusts an assembly a hundred meters down.

Sourcing underwater robot cable assemblies

Subsea harness sourcing rewards proven sealing capability over unit price. The harsh-environment case that opened this guide is the pattern: the customer expanded scope once the supplier proved it could deliver exact braided constructions, laser-etched marking, and injection-molded connectors. For underwater work, substitute 'overmolded, depth-tested, corrosion-planned' and the logic is identical — capability and evidence win the program.

Qualify for: overmolding and potting capability, wet-mate or sealed-penetrator experience, seawater-rated materials, and pressure/immersion test capacity. For prototype-to-volume, see our waterproof robot cable assembly and prototype cable assemblies; the IP framework is in our IP ratings guide.

When standard waterproofing is enough

Not every wet robot is a subsea robot. A pool-cleaning robot, a wash-down food-plant robot, or a splash-exposed AGV needs solid IP67/IP68 sealing and corrosion-aware materials, but not oil-filled, pressure-balanced umbilicals or wet-mate connectors. Over-engineering those into a shallow-water machine just adds cost.

Match the build to the actual depth and duty. Shallow or splash duty → robust waterproof assembly and good materials. True subsea depth, in-water mating, or long deployments → the full umbilical, sealing, and pressure-test discipline above. The dividing line is pressure and whether connectors ever mate underwater.

Key takeaways

• Wireless subsea comms is advancing but low-bandwidth; power and live video keep underwater robots on a tether.
• An ROV umbilical is a system: power + data (often fiber) + strength member + managed buoyancy, terminated into sealed penetrations.
• IP68 is not a depth rating; survive depth with overmolding, potting, water-blocking, and a pressure test.
• Wet-mate (in-water) vs dry-mate (in-air) connectors are not interchangeable; use wet-mate only when truly needed.
• Plan for galvanic corrosion and seawater-rated jackets; every exposed metal part is a corrosion candidate.
• Validate with IP, hipot, and pressure/immersion testing — a failed connector can cost a whole vehicle recovery.

References

• Remotely operated underwater vehicle (ROV), Wikipedia: ROV
• Autonomous underwater vehicle (AUV), Wikipedia: AUV
• Umbilical cable, Wikipedia: Umbilical cable
• IP Code (ingress protection), Wikipedia: IP Code
• Galvanic corrosion, Wikipedia: Galvanic corrosion
• Magnetoelectric effect (basis of BlueME antennas), Wikipedia: Magnetoelectric effect

Frequently asked questions

Do underwater robots use cables or wireless communication?

Both, for different jobs. Wireless subsea links (acoustic, optical, or magnetoelectric like BlueME) handle low-bandwidth telemetry between robots, but they top out around kilobits per second and cannot move serious power or live video. So remotely operated vehicles stay on a tethered umbilical that carries power and high-bandwidth data, while autonomous vehicles run on internal battery harnesses. The wiring remains the lifeline.

Is an IP68 connector safe for a deep underwater robot?

Not necessarily. IP68 means continuous immersion under manufacturer-specified conditions, often only about 1.5 meters — it is not a depth rating. At 100 meters an assembly sees roughly 11 bar of pressure driving water into every seam. For depth you need overmolding, potting, and water-blocking, validated by a hydrostatic pressure test to the working depth, not just an IP68 label.

What is an ROV umbilical and why can't it just be a normal cable?

An ROV umbilical is a complete assembly that bundles power conductors, data lines (often optical fiber), sometimes hydraulic lines, and a load-bearing strength member under a jacket tuned for buoyancy in water. The strength member carries the tow load so the conductors do not, and every element terminates into a sealed penetration. It is engineered as a system with its connectors and vehicle penetration, which is why it cannot be substituted with a plain length of cable.

What is the difference between wet-mate and dry-mate subsea connectors?

Dry-mate connectors are sealed and mated in air, on deck or in a sealed housing, and must never be connected underwater. Wet-mate connectors are designed to be mated and de-mated while submerged, excluding water from the contacts as they join, which is needed for in-water tool changes, subsea docking, and seabed reconfiguration. Wet-mate is much more expensive, so use it only where underwater mating is genuinely required.

How do you stop corrosion in underwater robot cable assemblies?

Use seawater-rated jackets (PUR/TPU families) instead of generic PVC, and plan dissimilar-metal junctions deliberately to limit galvanic corrosion at connector shells and clamps. Specify shielding and a defined ground return so the conductive seawater does not corrupt signal, and treat every exposed metal part as a corrosion candidate. For long deployments, add salt-spray/corrosion testing to the qualification.

How should an underwater robot cable assembly be tested before deployment?

Go beyond continuity. Require insulation resistance and hipot, an IP/immersion test, and a hydrostatic pressure test to the rated depth, ideally with the assembly powered and monitored during pressurization. Because a single flooded connector can force an expensive vehicle recovery, subsea buyers should fund validated, documented testing rather than the cheapest connector.

I need custom waterproof cable assemblies for an ROV/AUV. What should I expect?

Expect a capability-and-evidence sale, not a price sale. Qualify the supplier for overmolding and potting, wet-mate or sealed-penetrator experience, seawater-rated materials, and pressure/immersion test capacity. Plan a prototype phase with depth-rated validation before volume, and specify working depth, mating method (wet vs dry), and test acceptance on the drawing so quotes compare the same assembly.