How engineering is unlocking the secrets of the deep, and keeping divers safe

Engineering has enabled humans to achieve feats and access environments once beyond our natural reach. Cars have allowed us to travel faster and further than we could on foot; with planes we’ve overcome gravity and with rockets, we’ve escaped the atmosphere. However, despite covering roughly 71% of the surface of the planet and hosting most of its animal biomass, the ocean remains one of the least observed environments on Earth. 

Seawater is roughly 800 times denser than air, and pressure rises by one bar every 10 metres of depth, making it an inhospitable environment for humans. A combination of currents, reduced light penetration and sea floor sediment hamper visibility, while naturally occurring minerals and ions dissolved in seawater corrode metals and short-circuit electronics. To study this environment, engineers have helped to develop sonar-equipped ships, autonomous submersibles and satellites to help map the seafloor; and sensors that can track everything from temperature and salinity to the movements of animals such as whales to further increase our understanding of the ocean.

However, all these technologies map, count and track from a distance – comparable to studying the Amazon rainforest from a helicopter in the tree canopy, as opposed to on the forest floor. Direct observation without the barriers imposed by ships or submersibles requires immersion, which in practice, means diving. 

Over the last century, engineers have developed equipment that allows humans to interact with the ocean through diving. The most well-known of these are self‑contained underwater breathing apparatus (scuba) systems that let people carry an independent air supply on their backs.

These offer more freedom for divers than being tethered to a surface supply of air, or relying on holding their breath, as freedivers do. In its simplest form, an open‑circuit regulator (‘How does scuba gear work?’, Ingenia 102) delivers air from a high‑pressure cylinder on diver’s backs and vents exhaled gas into the sea. More advanced systems, such as rebreather units, capture and remove carbon dioxide from exhaled air using an absorbing material and add oxygen to the recycled gas, allowing for longer dives.

Inventions such as these have transformed safe interaction with the underwater world, and what divers can achieve, but they can’t overcome the physiological limits that present one of the biggest challenges in human underwater exploration regardless of how much breathing gas someone can carry with them – time spent underwater.

How long to stay underwater?

Mathematical tables developed in the early 1900s provide guidance on how long divers can spend at a given depth and still ascend safely. They highlight a trade-off between depth and time: the deeper you go, the less productive time you can spend underwater. Someone may be able to spend an hour at 15 metres with no stop on ascent; at 40 metres that window shrinks to about eight minutes.

Today, wrist-worn dive computers are used to simplify depth decisions in real time. Inside, a pressure transducer – often a piezoresistive strain gauge whose electrical resistance changes when squeezed – converts depth and pressure changes into an electrical signal. Coupled with a timer, a processor uses decompression models based on the same tables to display clear guidance on ascent rate, recommended pauses for safety stops and remaining no-stop time (the time you can remain at that depth without stopping on the way up).

After surfacing, the computer remembers the dive profile of the previous dive and uses it to adjust safety guidance for later dives, accounting for any remaining nitrogen that may still be leaving the body – useful when planning several dives over a short period. Staying within depth–time limits and following a dive computer’s real-time prompts keep the risk of decompression sickness (DCS) (see ‘Why slow ascents matter’) very low in recreational diving. 

Accessing deeper parts of the ocean, or staying for longer, introduces new challenges. Submersibles that can reach greater depths are expensive, and remotely operated vehicles separate people from the water, limiting direct interaction with the environment. When hands-on work must be done in deeper parts of the ocean – such as in subsea construction, research, or salvage operations – saturation diving is used. 

Saturation diving is widely used to support deep-water offshore projects in the oil, gas and wind industries through building, maintaining and inspecting subsea infrastructure. In this specialised technique, divers live at the same pressure as their underwater work site until the job is complete, either in an underwater habitat, or in pressurised chambers on the surface, and shuttle to the work site in a pressurised capsule called a diving bell.

Once the body has absorbed the maximum amount of gas at a given depth, it is ‘saturated’, and decompression time is fixed. Returning to the surface takes the same time whether the job lasts a day or two weeks. A mission at 200 metres requires roughly eight days to surface safely. Keeping the decompression to one event at the end of the mission increases the productivity of the time spent underwater and reduces the risk of DCS.

Accessing greater depths

Ocean technology company DEEP is developing underwater habitats to give scientists and researchers more time at depth – access once limited to specialist commercial saturation teams – in order to better understand the oceans. Safety is central to its work: the DNV, an independent maritime classing agency, verifies each stage of design, build and manufacturing operations. Phil Short, DEEP’s Underwater Research and Training lead explains that DEEP’s subsea habitats will be the first of their kind to be classified under DNV rules, adding that “we don’t mark our own homework”. 

Vanguard, the company’s pilot habitat, was unveiled in Miami in October 2025. It is designed to house four crew at a depth of about 15 metres for week-long missions. At this depth, the habitat can run on ordinary air mixtures and could boost the productivity of scientific missions such as coral restoration projects. DEEP repurposed a steel pressure vessel from a saturation system known as a surface decompression chamber, which would typically sit on the deck of a boat to keep saturation divers at pressure. A standard decompression chamber is an airtight vessel that provides a controlled, pressurised environment for divers to safely return to normal atmospheric pressure.

It is used both as a preventative measure before or after a dive and as a treatment for DCS. The chamber gradually reduces pressure, allowing excess dissolved gases such as nitrogen to safely exit the body and preventing the formation of dangerous bubbles in the bloodstream.

The engineers pressure tested the system to extreme internal and external depths and reengineered it into a comfortable living module classed by DNV. In-house engineers at DEEP designed a dive centre that is kept at the same pressure as the seabed to dock onto one end of the living vessel via a hatch, containing a moon pool – a floor opening to the ocean that allows entry and exit into the base. Extended time spent underwater in Vanguard will cause the crew to be saturated at the end of their mission. 

Vanguard sits on a foundation that anchors the habitat to the seabed. The habitat is connected to the surface via an umbilical that is attached to a large support buoy, which is roughly the size of two transit vans. The buoy houses compressors, generators with solar backups and systems that supply the habitat with breathing air, power and communications. If this buoy were to malfunction or become disconnected, Vanguard’s dive centre is ringed with multiple 55-litre gas cylinders that act as onboard reserves and hold enough gas for the crew to complete a full decompression and return to the surface safely.

Two other small independent habitats, known as refuges, are also fixed to the foundation and are accessible with their own moon pools to provide dry, safe environments for the crew to wait in if the main habitat cannot be used. Alongside supplies brought down in the umbilical from the surface support buoy, bladders – large, flexible containers that are commonly used within the commercial maritime industry – can be delivered to the habitat using submersibles or used for waste storage, which prevents anything from the habitat from being discharged into the ocean. 

To return, the connecting hatch that isolates the living module from the other compartments is shut so it doubles as an underwater decompression chamber. Valves slowly bleed gas over the course of a day to reduce its internal pressure. This gradual depressurisation allows the extra gas absorbed over the mission to leave the crew’s bodies until they are back at surface pressure. To leave, the living module is repressurised so the hatch to the dive centre can be opened.

Because the crew are no longer saturated from the week spent at sea, they can exit through the moon pool and ascend the same way a diver would after a 15-metre dive. Short explains that this decompressing approach, in which decompression is completed inside a shallow underwater habitat before a simple diving ascent, has been in use for over three decades in previous underwater habitat programmes run by NASA, the US Navy and scientific diving teams, but is only applicable for habitats at shallow depths.

Future subsea habitats that DEEP is working on aim to enable human deployments up to depths of 200 metres, and will build on the technology tested during Vanguard’s deployment.

These habitats will be modular, self-contained systems carrying their own life support, like a submarine, without the need for a surface support buoy. Components will be built using wire-arc additive manufacturing, a metal 3D-printing method that melts metal wire with a controlled electric arc at the tip of a welding torch and lays the metal down in layers to form the part. HexBot, DEEP’s six-arm robot, uses these torches to build curved pressure-rated segments up to 6.2 metres in diameter. This process has DNV approval and Short points out that the approach lets them print, inspect and pressure-test individual components, rather than commit to a full hull upfront. In tackling unsolved problems, DEEP recognised the limiting factor was usually “[the lack of] enough budget to enable the engineering community to solve the problem”, rather than if it was doable. 

Short draws on 35 years of diving and teaching experiences across the globe to help lead the development of dive systems and training procedures for DEEP’s future aquanauts who will live in its habitats. He says the goal is to “merge the safest bits of multiple dive communities” and create a programme that will prepare everyone from experienced commercial divers to non-diving scientists. Using astronaut training as a comparison he says “in the early space age, only elite military test pilots with thousands of flying hours went up. Now researchers [go] as they have an extreme scientific need to be there. We aim to do exactly the same [for the sea].” As Short puts it, “you can’t bypass the limits” but DEEP is an example of using engineering to make “systems that safely work within them”.