Some challenges in subsea operations remain genuinely unsolved — not merely neglected, but actively difficult despite significant research and engineering effort. This page surveys the most significant open problems and why they matter for operational safety, data quality, and regulatory compliance.

Why This Exists

Understanding that a problem is genuinely open (rather than just unimplemented) helps operators set realistic expectations, avoid overconfidence in inadequate solutions, and direct research and development investment appropriately.

Who This Is For

  • Engineers and researchers working at the frontier of subsea systems
  • Standards body participants scoping future standards work
  • Operators evaluating emerging technology claims
  • Project managers setting realistic capability expectations

The Problem

Precise subsea positioning without pre-deployed infrastructure (LBL transponders, seabed beacons) remains unsolved at scale. USBL accuracy degrades with depth. INS/DVL systems accumulate error over time. GPS is unavailable underwater.

Current State

Best available solutions combine INS + DVL + periodic USBL aiding, achieving decimetre accuracy over kilometre-scale missions in good conditions. In challenging terrain or poor acoustic conditions, errors can be metres to tens of metres.

Why It Matters

  • AUVs operating in unmapped terrain cannot verify their own position
  • Autonomous collision avoidance requires reliable position knowledge
  • Survey data attributed to incorrect positions is misleading

Active Research Directions

  • Terrain-relative navigation using sonar maps
  • Cooperative navigation between multiple vehicles
  • Improved acoustic methods for longer range/depth

Long-Duration Energy Storage

The Problem

Underwater vehicles are energy-limited. Batteries have fixed capacity, and recharging requires surfacing or docking. Long-duration autonomous operations (days to weeks) require either very large batteries, slow speeds, or underwater recharging.

Current State

Lithium-based batteries provide energy density of ~200 Wh/kg. Fuel cells offer higher density but add complexity and hydrogen storage challenges. Underwater recharging works for vehicles that can return to a dock, but remains unreliable for extended range operations.

Why It Matters

  • Mission duration limits what can be monitored or inspected autonomously
  • Return-to-dock requirements constrain operational area
  • Energy availability affects emergency response capability (a vehicle that cannot abort safely is a hazard)

Reliable Subsea Communication

The Problem

Acoustic communication provides the only long-range data link for subsea vehicles, but bandwidth is severely limited (typically <10 kbps) and latency is high (seconds per exchange). Optical communication provides high bandwidth but over very short ranges (<100m in clear water).

Current State

No technology provides both long-range and high-bandwidth subsea communication. Operations requiring rich real-time data exchange must use short umbilicals or accept very low data rates.

Why It Matters

  • Autonomous vehicles that cannot communicate cannot receive updated mission plans or abort commands
  • Real-time monitoring of complex situations requires data rates that acoustic links cannot provide
  • Swarm coordination at range is limited by available communication bandwidth

Fault Tolerance in Autonomous Decisions

The Problem

How should an autonomous vehicle behave when it encounters a situation outside its design envelope? Sensors may fail partially (giving wrong data rather than no data). Environmental conditions may be outside tested ranges. Multiple faults may interact unexpectedly.

Current State

Most autonomous systems handle individual, anticipated faults. Combinations of faults, especially those involving sensor data that is wrong rather than missing, are much harder. There is no general solution for safe behaviour under arbitrary fault combinations.

Why It Matters

  • Vehicles that behave dangerously under unexpected faults cannot be safely deployed in complex environments
  • Regulatory certification of autonomous safety behaviours requires confidence in fault response
  • Incident investigation requires understanding what the system “thought” it was doing

Standardised Autonomy Certification

The Problem

There is no accepted standard for certifying that an autonomous subsea system is safe to operate. Aviation (DO-178C for software, DO-254 for hardware) and automotive (ISO 26262) have developed rigorous frameworks over decades. Subsea autonomy has none.

Current State

Certifiers apply analogous standards by judgement. Different jurisdictions require different approaches. Novel systems face multi-year regulatory uncertainty.

Why It Matters

  • Operators cannot demonstrate regulatory compliance for autonomous deployments
  • Liability is unclear when an autonomous system causes harm
  • Innovation is slowed by regulatory uncertainty

Subsea Environmental Interaction Prediction

The Problem

Predicting how a vehicle will interact with the environment — particularly in near-seabed operations, strong currents, and turbid water — is difficult. Vehicle dynamics models built in controlled conditions do not always transfer to field conditions.

Why It Matters

  • Collision avoidance near seabed structures requires accurate motion prediction
  • Safe operation near sensitive environmental sites requires knowing the vehicle’s footprint
  • Survey quality depends on maintaining accurate track spacing