Of the three domains covered in this series — aerial, maritime, and subsea — subsea is the most unforgiving. A UAV that loses power glides. A USV that loses power drifts. An AUV that loses power sinks, and the ocean provides no second chances for a vehicle that reaches the bottom at operating depth. Every engineering decision in subsea systems design is made against this backdrop: the consequence of failure is not mission loss but vehicle loss, and in deep water, permanent loss. The engineering disciplines that matter most in subsea design — pressure hull integrity, sealing, buoyancy management, and energy budget — are disciplines that have no direct equivalents in aerial or maritime engineering, and where the margins of error are measured in bars of pressure rather than meters of altitude.
The subsea design challenge
The fundamental challenge of subsea engineering is that water pressure increases by approximately 1 bar (14.7 psi) for every 10 meters of depth. At 100 m depth — a relatively shallow operational range for a tactical AUV — the ambient pressure is 11 bar absolute, or approximately 160 psi. Every enclosure, every seal, every cable penetration, and every structural joint must withstand this pressure without deformation, leakage, or failure for the duration of the mission. Unlike structural loads in aerial systems — which are transient and associated with specific maneuvers — hydrostatic pressure is continuous, acts on all surfaces simultaneously, and increases linearly with depth. There is no altitude relief valve, no altitude hold mode that reduces the load. The structure is under full operating pressure from the moment it submerges until the moment it surfaces.
The second fundamental challenge is the absence of electromagnetic communication through water. Radio frequency signals attenuate extremely rapidly in seawater — a 100 MHz signal attenuates by approximately 300 dB per meter in seawater, making any RF-based communication or navigation system completely useless below the surface. GPS, which is the foundation of aerial and maritime navigation, provides exactly zero position information to a submerged vehicle. The entire navigation and communication architecture of a subsea system must be built on acoustic signals, tethers, or pre-programmed autonomy — and each of these solutions has limitations that drive specific design constraints.
Pressure hull design
The pressure hull is the primary structural element of any AUV — the sealed vessel that contains the electronics, power, and payload at near-atmospheric pressure while the ambient environment is at tens or hundreds of bars. Its design is governed by the buckling and collapse pressure of the chosen geometry and material, not by the yield strength of the material alone.
Cylindrical pressure vessels — the standard form for AUV hulls — fail primarily by buckling rather than by material yield under external pressure. Buckling is a geometric instability: the cylinder collapses inward at a critical pressure that depends on the ratio of wall thickness to diameter (t/D), the cylinder's length-to-diameter ratio (L/D), and the material's elastic modulus. Thin-walled cylinders with large L/D ratios buckle at pressures well below the material's yield strength. The critical external pressure for a long thin-walled cylinder is:
Material selection for the pressure hull trades weight, corrosion resistance, machinability, and pressure rating. Aluminum 6061-T6 is the most common choice for AUV hulls at moderate depths (to 300 m): it machines cleanly, is available in standard tube and bar stock, provides a good strength-to-weight ratio, and is adequately corrosion resistant with anodizing and sacrificial anode protection. For deeper applications, titanium alloys (Ti-6Al-4V) offer higher specific strength at greater cost. Carbon fiber reinforced polymer is structurally attractive but requires careful design to avoid the inter-laminar shear failures that external hydrostatic pressure can drive — the laminate must be oriented to resist the radial compressive loading without delamination, and any through-holes for O-ring grooves or fasteners create stress concentrations that are harder to manage in composites than in isotropic metals. The AUV/UUV hull used SLS PA-12 Nylon with a Kevlar overwrap — a design choice driven by the modularity requirement, the available manufacturing process, and the relatively shallow operational depth target, where the pressure demands were manageable within the polymer's structural capability.
O-ring sealing and end cap design
The O-ring is the fundamental sealing element in AUV design. Every access port, every end cap, every cable penetration, and every viewport requires at least one O-ring to maintain the pressure boundary. Understanding O-ring sealing mechanics — not just specifying a standard groove and an off-the-shelf ring — is essential for a hull design that is reliable over many dive cycles.
An O-ring seals by being compressed into a groove, deforming to fill the gap between the mating surfaces. The compression is defined as the percentage reduction in the O-ring's cross-sectional diameter: (free diameter − installed height) / free diameter. For static face seals on end caps, the standard compression range is 15–30%. Too little compression provides an inadequate seal — the ring does not conform to minor surface irregularities. Too much compression over-stresses the elastomer, accelerates stress relaxation, and can cause the O-ring to extrude into the clearance gap between the mating parts under pressure, damaging the ring and causing catastrophic seal failure.
Under external hydrostatic pressure, the pressure itself aids the seal — the external pressure forces the O-ring into tighter contact with the groove walls, increasing the contact stress with depth. This is why a properly designed AUV O-ring seal actually becomes more reliable with increasing depth, as long as the O-ring material, groove geometry, and surface finish are correctly specified. The failure mode is typically not at maximum depth — it is at shallow depth on the way down, where the pressure has not yet developed to press the O-ring into full contact, and where a minor surface imperfection in the groove or a nick in the O-ring can allow a small flow path to develop before the pressure closes it.
Buoyancy and depth control
Unlike a UAV that generates lift through wing geometry and a USV that sits at the surface by inherent buoyancy, an AUV must actively manage its position in the water column. The three buoyancy states are positive (the vehicle is lighter than the water it displaces and rises to the surface), negative (the vehicle is heavier than the water it displaces and sinks), and neutral (the vehicle is exactly in equilibrium and neither rises nor sinks). Most AUVs are designed to be slightly positively buoyant — if propulsion fails, the vehicle rises to the surface rather than sinking to the bottom. This is the subsea equivalent of the "glide to safety" design philosophy in aircraft, and it is the primary reason that AUV loss to propulsion failure, while expensive, does not necessarily mean permanent loss.
The vehicle's buoyancy changes with depth because the hull compresses slightly under hydrostatic pressure, reducing its displaced volume, while the mass remains constant. This compressibility effect means a vehicle that is neutrally buoyant at the surface will become slightly negatively buoyant at depth — a well-known trap for designers who set buoyancy only at the surface. For aluminum hulls, the compressibility is small but measurable. For composite hulls with higher compliance, the depth-dependent buoyancy shift can be significant. The design must account for the buoyancy state at the maximum operational depth, not just at the surface.
Variable buoyancy systems (VBS) — oil bladder systems that change the vehicle's displaced volume by pumping oil between an internal reservoir and an external flexible bladder — allow the AUV to control its depth without using propulsion thrust. A vehicle with a VBS can hover at depth, glide up or down by adjusting buoyancy and using the vehicle's wings or fins to convert vertical motion to horizontal flight, or park on the bottom and wait. The AUV/UUV design achieved its 110-hour battery life and 275-nautical-mile range in part through gliding efficiency: propulsion was used for transit at speed, while depth adjustments during slower survey passes used the VBS rather than continuous thrusting, dramatically reducing power consumption during those phases.
Underwater propulsion
AUV propulsion encompasses more than the ducted thrusters that dominate the commercial and academic markets. The propulsion choice — like every other design decision in this series — flows from the mission requirements: the speed required, the depth rating, the acoustic signature permitted, and the maneuverability demanded by the mission profile.
Ducted thrusters enclose the propeller in a duct or shroud, improving efficiency at low advance ratios (high thrust demand at low speed), reducing tip vortex losses, and protecting the propeller from debris and the seafloor. Duct material for seawater exposure is typically glass-reinforced nylon or HDPE. Multi-thruster 6-DOF configurations — longitudinal thruster for surge, lateral thrusters for sway, vertical thrusters for heave — provide full hover and translation capability at the cost of hydrodynamic drag from the multiple thruster housings and their interactions with the flow field. 6-DOF is appropriate for inspection, manipulation, and confined-space operations where maneuverability matters more than efficiency.
Single open-screw propulsion — the classic torpedo configuration — produces higher efficiency than ducted thrusters at moderate to high advance ratios, where the duct's benefit at low speed becomes a drag penalty at cruise speed. The single aft screw drives a persistent torque reaction that the vehicle's fins must counteract, producing a small but continuous hydrodynamic drag penalty. It is mechanically simple, pressure-compensated through an oil-filled shaft seal, and well-proven across decades of heavyweight torpedo and AUV programs.
Counter-rotating propellers
The counter-rotating propeller (CRP) system uses two coaxial propellers on the same shaft axis, rotating in opposite directions. The forward propeller extracts energy from the incoming flow and discharges a rotating wake. The aft propeller, rotating in the opposite direction, recovers the rotational kinetic energy in that wake — energy that a single-screw design leaves in the water as wasted swirl. The efficiency gain over a single-screw design of equivalent diameter and pitch is typically 10–15%, depending on the propeller geometry and the operating point.
The second advantage of CRP is torque cancellation. A single-screw propulsion system generates a net torque on the vehicle equal and opposite to the motor torque. The vehicle's control surfaces must counteract this torque to maintain straight-line heading — a continuous efficiency penalty that increases with motor power. A CRP system, with the two propellers producing equal and opposite torques, cancels this reaction at the propeller axis. The vehicle experiences no net roll or yaw moment from the propulsion system, simplifying the control law and eliminating the fin deflection and associated drag required to compensate for a single-screw torque offset.
The AUV/UUV design used a submarine-style dual counter-rotating propeller system for main longitudinal propulsion, rather than the ducted thruster arrays common in commercial inspection AUVs. This choice reflects the mission profile: a vehicle optimized for long-range transit and endurance at moderate speed, where the CRP efficiency advantage compounds over a 110-hour mission and the torque cancellation eliminates a source of trim drag that would accumulate over 275 nautical miles. Additional thrusters were provided for station-keeping and low-speed maneuvering, but the primary propulsion — the system that consumes most of the energy budget — was the CRP.
Magnetohydrodynamic propulsion
Magnetohydrodynamic (MHD) propulsion — the "caterpillar drive" of the Hunt for Red October — is the only propulsion system in AUV engineering that contains no moving mechanical parts. Its operating principle is the Lorentz force: when an electrical current is passed through a conductive fluid in a magnetic field, the fluid experiences a force perpendicular to both the current and the magnetic field. In an MHD thruster, seawater flows through a channel containing powerful magnets and electrode arrays. Current is injected through the seawater perpendicular to the magnetic field; the Lorentz force accelerates the seawater through the channel and out the nozzle, producing thrust.
The physics is elegant. The engineering challenge is the electrical conductivity of seawater, which is approximately 5 S/m — far lower than any metallic conductor. To generate meaningful thrust through a modest-diameter channel, the product of the magnetic field strength (B) and the current density (J) must be large enough that the resulting body force on the water exceeds the viscous and pressure drag of the channel. With conventional electromagnets producing fields of 1–2 Tesla, the current densities required for useful thrust produce significant ohmic heating in the seawater, consuming power without producing thrust. The Yamato-1 experimental ship — the only full-scale MHD vessel ever operated — demonstrated this in 1992, reaching approximately 8 knots at an overall propulsive efficiency below 0.5%. The electrical energy consumed to produce the thrust was dominated by resistive losses in the seawater, not by useful thrust work.
The case for MHD propulsion is not efficiency — it will not compete with a well-designed propeller system on energy-per-Newton-of-thrust. The case is acoustic stealth. A propeller, even a well-designed low-noise one, produces blade-rate tonal noise, shaft bearing noise, and cavitation noise. An MHD thruster produces none of these. The only acoustic signatures are the flow noise of water moving through the channel and the very low-frequency electromagnetic signature of the current loops — both of which are orders of magnitude quieter than any mechanical propulsion system at equivalent thrust levels. For a vehicle operating in an environment where acoustic detection by passive sonar is the primary threat, this trade — lower efficiency for dramatically lower acoustic signature — may be exactly the right one to make.
Recent testing has confirmed that at small scales, with careful magnetic circuit design and optimized electrode geometry to minimize ohmic losses, MHD produces thrust levels that are operationally useful for low-speed maneuvering and station-keeping. The efficiency gap with mechanical propulsion remains substantial, but the gap is narrowing as high-energy permanent magnet technology advances. A hybrid architecture — CRP or single screw for transit, MHD for the final approach and station-keeping phases where acoustic signature is most critical — represents a credible engineering solution for programs where stealth is a primary requirement alongside endurance.
Thruster motors for subsea applications are either oil-compensated (motor stator filled with non-conductive oil at pressure, eliminating pressure differential across the housing — works at any depth, more complex sealing) or pressure-tolerant (motor operates with water in contact with stator windings using seawater-compatible magnet wire insulation — simpler but rated to specific depth limits).
Remotely Operated Vehicles
An AUV and an ROV are both unmanned subsea vehicles, but they are answers to different engineering questions. An AUV asks: how do we operate for long duration over a large area without a human in the loop? An ROV asks: how do we put a human's eyes, hands, and judgment at depth in real time, without putting a human body there? The tether is not merely a design feature of the ROV — it is the architectural decision from which everything else follows, and it is the reason that ROVs and AUVs are complementary systems rather than competing ones.
The tether provides three things that an AUV must carry internally at great cost in weight and volume: power, high-bandwidth communication, and a continuous mechanical connection to the surface. A work-class ROV drawing 50–200 kW of electrical power through its umbilical can run powerful thrusters, high-intensity lighting, HD cameras, sonar systems, and manipulator arms simultaneously — a power budget that would require a battery system of impractical weight and volume for any AUV. The live HD video feed and real-time joystick control that the tether enables allow a skilled pilot to perform precise manipulator work, sample collection, and structural inspection tasks that no autonomous system can reliably replicate at current technology levels.
ROV classes and their engineering implications
ROVs are typically classified by their capability and physical scale into three broad classes, each with distinct design requirements.
Observation class ROVs are small, lightweight vehicles — typically under 20 kg in air — designed for inspection, survey, and reconnaissance tasks that do not require physical intervention. They carry video cameras, lights, and sometimes a small sonar, but no manipulators. Their thrusters are sized for observation-speed maneuvering rather than station-keeping against strong currents. The open-frame structure is minimal, and the umbilical is thin (diameter 10–15 mm) and positively buoyant. These are the ROVs used for pipeline inspection, hull surveys, and scientific observation. Their small size and modest power demand make them deployable from small vessels or even by a single operator from a dock.
Work-class ROVs are large vehicles — typically 1,000–5,000 kg in air — designed for active intervention: operating valves, connecting and disconnecting equipment, cutting and welding, retrieving objects, and performing tasks that require significant force. They carry one or two manipulator arms with multiple degrees of freedom, high-power thrusters capable of maintaining position against significant currents while the manipulators apply force to structures, powerful lighting, multiple cameras, and often a full sonar suite including multibeam imaging sonar. Their umbilicals are thick (50–80 mm diameter) fiber-optic composite cables carrying hundreds of kilowatts of electrical power and multiple high-bandwidth data channels. Work-class ROV deployment requires a dedicated launch and recovery system (LARS) and a support vessel large enough to handle the winch and the vehicle's weight.
Micro and mini ROVs sit below observation class — hand-portable vehicles of 3–15 kg with compact thruster configurations, suitable for harbor inspection, confined-space entry (flooded compartments, culverts, pipelines), and rapid deployment from any accessible water surface. Their umbilicals are typically neutrally buoyant, thin cables 8–10 mm in diameter carrying power and a composite video signal or compressed digital video over a twisted pair.
Open-frame construction
Where an AUV uses a streamlined pressure hull to minimize hydrodynamic drag during high-speed transit, an ROV uses an open frame — a structural skeleton of aluminum or HDPE tube, extrusion, or plate — because ROV operations are almost always conducted at low speed in a specific area rather than in long-range transit. The open frame is not a compromise; it is the correct structural solution for the ROV's operating envelope.
The open frame provides unrestricted access to the electronics, thruster modules, manipulators, and payload — critical for a vehicle that must be regularly serviced and reconfigured between dive operations. Components are bolt-mounted to the frame and replaceable at the surface without disassembling the vehicle. The electronics are housed in individual pressure housings bolted to the frame — one housing for the main electronics, one for the camera systems, one for the sonar — each independently pressure-rated and independently removable. This modularity is the practical key to ROV reliability: a failed electronic housing is swapped for a spare in 20 minutes; a failed AUV pressure hull may require complete disassembly.
Frame material for seawater exposure is typically 6061 aluminum alloy with hard anodizing, or HDPE for components where galvanic corrosion is a concern. Stainless steel fasteners with nylon inserts are the standard fastening system — the nylon insert breaks the galvanic circuit between the stainless steel fastener and the aluminum frame, preventing the accelerated corrosion that occurs when these two materials are in direct contact in seawater.
The umbilical: engineering the tether
The umbilical is the most mechanically stressed component on any ROV system, and it is the source of more operational problems than any other single element. It must carry electrical power (often at high voltage to minimize resistive losses over long tether lengths), high-bandwidth fiber-optic data, and the mechanical tension of supporting the vehicle and the tether's own weight in the water column — all while being flexible enough to spool on a winch drum, tough enough to survive contact with sharp edges on subsea structures, and buoyant enough not to drag the vehicle down when deployed in a catenary.
Electrical power is typically transmitted at 3,000–5,000 V for work-class systems, stepped down to the vehicle's operating voltage by a transformer inside the vehicle's electronics housing. High transmission voltage is necessary to minimize the I²R resistive losses in the power conductors over tether lengths of 300–1,000 m — a 50 kW ROV running on 48 V at the vehicle would draw approximately 1,000 A, and the resistive losses in even a short tether at that current would make the system impractical. At 3,000 V, the same power requires only 17 A, and the resistive loss in a 300 m tether with reasonable conductor cross-section is manageable.
Fiber-optic cores in the umbilical carry real-time HD video, vehicle telemetry, sonar data, and control signals. A modern work-class ROV tether carries 8–12 single-mode fiber cores, providing several gigabits per second of aggregate bandwidth — enough for multiple simultaneous HD video channels and all sensor data streams with significant headroom. The fiber cores are fragile relative to the electrical conductors and the strength member, and must be protected from crush and bend loads by a loose-tube buffer inside the cable structure.
ROV thruster configuration and station-keeping
ROV thruster configurations are designed for 6-DOF control — the vehicle must be able to move in any direction and rotate about any axis independently, because work tasks require precise positioning of the vehicle relative to a structure while the manipulator applies force. A typical work-class ROV uses four horizontal thrusters (arranged as two pairs, each pair in a cross configuration that provides surge, sway, and yaw control) and two or four vertical thrusters for heave control. The horizontal thrusters are often tilted at 45 degrees to the vehicle's lateral and longitudinal axes, which provides both surge and sway thrust from each thruster and eliminates the need to switch between different thruster groups for different directions of motion.
Station-keeping — maintaining position against current while the manipulator applies force to a structure — is the most demanding operating condition for the thruster system. The thruster must provide enough thrust to hold the vehicle stationary against the combined drag of the current on the vehicle body and the reaction force from the manipulator's work. A work-class ROV operating a valve in a 1-knot current may experience current drag of 200–400 N on the vehicle body, plus a manipulator reaction force of 500–1,000 N depending on the task — requiring 700–1,400 N of total sustained thrust from the horizontal thruster array.
AUV vs. ROV: the engineering decision
The choice between an AUV and an ROV is the subsea equivalent of the Module 02 configuration trade study — a decision that must be made from the mission requirements, not from familiarity with a particular technology. The key questions are: does the mission require real-time human judgment and physical intervention (ROV), or does it require large-area coverage and long-duration autonomy (AUV)? Can a tether be managed safely in the operating environment (ROV), or does the operating area, depth, or current regime make tether management impractical (AUV)?
Subsea power systems
Energy storage for AUVs faces constraints that do not apply to aerial or maritime applications. The battery must be contained within the pressure hull or in a pressure-tolerant enclosure — it cannot simply be bolted to a frame in open air. Battery chemistry, cell format, and management system must all be compatible with the sealed, pressure-compensated environment.
Li-Ion cells in cylindrical format (18650, 21700) are the standard for AUV energy storage. They are available in very high energy density configurations (>250 Wh/kg at the cell level), package efficiently into cylindrical battery packs that fit within circular cross-section AUV hulls, and have well-characterized behavior under pressure. The battery management system (BMS) must be housed within the pressure hull and sized to monitor all individual cells for voltage, temperature, and state of charge — a cell-level fault in a tightly packed AUV battery pack can produce a thermal runaway event in a sealed environment with nowhere for the heat and gas to go. The consequences of a thermal runaway event inside a pressure hull at depth are catastrophic. Cell selection, BMS design, and charging procedures are more critical in an AUV than in almost any other battery-powered application.
The 110-hour endurance of the AUV/UUV design required a very large energy storage system — approximately 25–30 kWh depending on the mission profile — packed into the hull volume available within the weight budget. This drove the selection of the highest energy density Li-Ion cells commercially available, a battery management system capable of monitoring over 100 individual cells, and a charging protocol that was conservative to avoid degrading cell capacity over the vehicle's operational life of several hundred dives.
Navigation without GPS
GPS provides no position information to a submerged vehicle. Underwater navigation is built on four principles: acoustic positioning, inertial navigation, Doppler velocity measurement, and terrain-referenced navigation. In practice, AUV navigation systems combine two or more of these to achieve the position accuracy required by the mission.
Doppler Velocity Log (DVL) is the most widely used underwater velocity reference. The DVL transmits acoustic pulses toward the seafloor at oblique angles and measures the Doppler shift of the reflected signals to determine the vehicle's velocity over ground in three axes. DVL provides accurate velocity (typically ±0.1% of speed) but no absolute position — it must be integrated with inertial navigation to produce a position estimate. DVL-aided INS systems accumulate position error at approximately 0.1–0.5% of distance traveled, producing position errors of 1–5 m per kilometer — sufficient for most survey tasks but insufficient for precision operations near fixed infrastructure.
Ultra-Short Baseline (USBL) positioning uses a transducer on a surface support vessel or a fixed seafloor transponder to measure the range and bearing to the AUV, providing an absolute position reference that corrects the accumulated INS drift. USBL accuracy is typically 1–3 m at ranges up to 1,000 m, degrading with range and with the angular accuracy of the shipboard system. The limitation is that USBL requires a surface asset in acoustic contact with the AUV — it is a tethered positioning system in the acoustic sense, even when the vehicle is physically untethered.
Long Baseline (LBL) positioning deploys a network of seafloor transponders at known positions. The AUV measures its range to each transponder and computes its position by triangulation. LBL provides absolute position accuracy of 1–3 m across a large area without requiring a surface asset overhead, but requires the initial deployment and survey of the transponder network, which is time-consuming and expensive. LBL is appropriate for high-value, long-duration programs where the setup cost is amortized over many missions.
Acoustic communication
Acoustic modems are the primary communication link between a submerged AUV and the surface or a support vessel. Unlike radio, acoustic signals propagate efficiently through water — but with severe bandwidth limitations. A high-performance acoustic modem achieves 10–40 kbps at ranges up to 1,000 m, compared to megabits per second for even a modest RF link. At longer ranges (5–10 km), bandwidth drops to 1–5 kbps. This is sufficient for vehicle health telemetry, position reporting, and simple command uplink — but it is nowhere near sufficient for real-time video or high-resolution sensor data streaming.
The acoustic channel is also subject to multipath — reflections from the surface, the seafloor, and thermal layers in the water column that cause the signal to arrive at the receiver via multiple paths with different delays, corrupting the original signal. Acoustic modems use sophisticated signal processing to resolve these multipath arrivals, but the processing overhead limits the achievable bit rate and introduces latency. The round-trip delay for an acoustic command at 1,000 m range — 0.67 seconds for the ping, plus 0.67 seconds for the acknowledgment — means that any command sent to the AUV takes at minimum 1.34 seconds to be confirmed received. Closed-loop control of an AUV through an acoustic modem is impractical; the vehicle must be sufficiently autonomous to execute its mission without real-time human control, using the acoustic link only for monitoring and high-level command updates.
The practical consequence for AUV mission design is that the vehicle must be capable of completing its primary mission entirely autonomously, navigating to within a pre-planned waypoint sequence and returning to the recovery point without any acoustic intervention. The acoustic modem provides mission monitoring and abort capability, but the mission design must not depend on acoustic commands for safety-critical actions. The AUV/UUV design operated on a purely pre-programmed mission for the deep survey phases, using surface pop-up events for position correction and acoustic check-in rather than continuous acoustic guidance.
Deployment, recovery, and mission planning
An AUV that cannot be reliably deployed and recovered is not an operational system, regardless of how capable it is underwater. The deployment and recovery system is determined by the available support infrastructure, the sea state limits of the operation, and the AUV's physical characteristics — weight, size, and the presence or absence of recovery aids such as acoustic pingers, flashing lights, and flotation systems.
For a small AUV (under 50 kg), hand deployment from a rigid inflatable boat in sea state 2–3 is practical. The AUV is powered up, mission loaded, and GPS-fixed on the deck before deployment. Deployment is a one-person lift and slide over the transom — no special handling equipment required. Recovery in the same sea state requires the vehicle to surface within visual range of the support boat, ideally guided by the acoustic modem or by a radio-frequency homing signal (permitted above the surface). The recovery point should be a minimum of 500 m downwind and downcurrent from the surface pop-up position to allow the boat to approach safely without running over the vehicle.
Mission planning for AUV operations must account for the ocean current field in the operational area. Unlike a UAV that can compensate for wind by adjusting its heading, an AUV operating near neutral buoyancy in a current is carried by that current while its DVL velocity reference measures velocity relative to the seafloor, not relative to the water mass. If the mission requires a specific seafloor coverage pattern, the waypoint sequence must be planned in seafloor-referenced coordinates, accounting for the fact that the vehicle's absolute geographic track will differ from the waypoint sequence by the integral of the current velocity over the mission duration.