Extreme (and Not So Extreme) Multistage Sidemount Cave Diving with AI

Photo courtesy of Natalie Gibb | Under the Jungle.

Soft flakes of microbial goo slapped my face as I scootered through them at high speed on my DPV. My exploration partner Vincent Rouquette-Cathala and I were back on our DPVs at our massive exploration project, Pandora, returning from an end of the line push dive. The ninety-minute scooter ride out was actually the most difficult part of the dive.

On the way into the cave to explore, the visibility is reduced from moderate concentrations of hydrogen sulfide in the water, but it's still possible to see about ten meters. On the way in, we weave our DPVs through narrow, dark tunnels like a slalom course, sometimes ducking our heads behind the props of the scooters to avoid hitting the ceiling. With an average depth of 24 meters/80 feet and dive times reaching between 5-6 hours, it's essential to make good time in and out of the cave to reduce decompression obligations in the adverse environment.

The main difficulty of dive exploring Pandora, however, is not the hydrogen sulfide, the tight tunnels that we must take at a high speed, or even the fact we have to drag four or five tanks and two scooters each through restricted passages, it's the presence of microbial growth covering nearly every surface of the cave. Our movement through the water disturbs the microbes, and flakes of microbial life rain down as we move through the tunnels, it remains suspended in the water for days. On the way out of the cave, while the visibility is technically alright, the flakes reduce your ability to see more than a meter or so ahead.

The scooter ride out is harrowing, requiring extreme concentration. You cannot look away while you are moving forward, not for even a split second. Stopping to make a gear adjustment or check a pressure gauge only further reduces the visibility, and it is essential from a safety standpoint to keep moving so that the water stays somewhat clear for you partner behind you. You can't stop and you can't look down for even a moment or you risk hitting the ceiling and damage your gear and your head.

Photo courtesy of Natalie Gibb | Under the Jungle.

Shearwater's transmitters have greatly improved my safety during exploration. I can slide my computer hand just up into my peripheral vision and determine when it's time to switch from my right stage to my left stage bottle without breaking visual contact with the environment. My gas management stays on point, and I don't run into the ceiling or a wall because I looked down for a split second to check an SPG. In this sort of environment, AI is indisputably safer.

As a full-time cave diving instructor and dive shop owner, I have been using Shearwater AI every day, even on normal “easy” cave dives, and I feel that it has improved my safety and awareness both as a diver and as an instructor. There's a lot to unpack and consider when switching to AI, especially on big dives with multiple stages.

Is AI Accepted in Cave Diving?

When I began brainstorming this article, I received a timely message from an Instagram follower. He wrote, “Has AI generally been accepted for 'light caving'? I know some people that are opposed to it due to potential failures that wouldn't necessarily be there with an SPG.”

This made me think. First of all, what is 'light caving'? Is there even such a thing? As a cave diving instructor who teaches everything from beginner cavern diving courses all the way through multi-stage and DPV cave specialities, I know there is a difference in penetration, depth, and dive time as my students progress to longer and more technically complex dives. However, I would never consider any foray into the overhead environment “light.”

Cave diving is cave diving, and the moment a diver enters the overhead environment, his or her gear, skills, and techniques must necessarily be top notch. There's no gear that I use during cavern training that I wouldn't use on a more extreme cave dive. When it comes down to it, there's a rock ceiling over your head, there's no way to ascend immediately, and you must plan accordingly. If I find value in transmitters and AI for multi-stage DPV cave exploration, I find the same value for cavern diving. If I accept a tool for easy cave dives, it would be hypocritical to deny its use on longer and more intense ones.

The real question here, is AI accepted in cave diving? Period. Ten years ago it certainly wasn't, and with good reason, as it wasn't uncommon to lose the signal from your transmitter. However, Shearwater's innovations and software, and specifically the SWIFT transmitters, make losing communications (comms) between the transmitter and your computer unlikely. It's not a problem I have had with the SWIFTs.

There are certainly cave divers and technical divers that frown upon AI technology, but I feel that cave diving is still a sport in its infancy. It's continually evolving and in the fifteen years that I have been working in cave diving I have observed initial resistance to many innovations that are now widely accepted. Handheld (without a cable or canister) primary lights were once laughed at, sidemount was considered risky and unstandardized, and even trimix was once thought to be dangerous and experimental. Yet all of these innovations have now become mainstream in the sport. I feel that cave diving should innovate and evolve. We should use technology to our benefit, and I am excited to be part of the newer generation embracing AI technology.

Is AI accepted in cave diving? It is at my shop and in my training programs. There's some huge benefits to it, both from a diver and and instructor standpoint. My entire guiding and instructor staff now use Shearwater AI and SWIFT transmitters. Before getting to the benefits of AI for sidemount and multi-stage sidemount, let's look at where things can go wrong.

SPG vs Transmitter Failures

My timely Instagram inquisitor asked about failures that were possible with transmitters that are not possible with SPGs. It's true, there are definitely failure points in transmitters that are not present with SPGs. However, few people in the “anti-transmitter” camp stop to think about the failure points that are present with SPGs that are not inherent with transmitters. There's quite a few, and based on the failures I have experienced with SPG use, I feel that using transmitters actually reduces the potential failure points on my regulators.

Failures of transmitters would be limited to a few items; as transmitters are basically very fancy and very expensive port plugs. The failures I would consider are, for the most part, avoidable through proper gear checks, and are listed below:

1. Loss of transmitter battery: if your transmitter battery was to run out of power, you would no longer be able to read your tank pressure(s). The CR2 batteries used in Shearwater's transmitters last for hundreds of dives. I spend nearly four hours underwater almost every day, and I replace my batteries about once every six months, before replacement is needed. In any event, a low battery would not be a surprise, as the transmitters send a “low batt” warning to the computer, and then continue to function for several more dives. You have time to switch the batteries out, and failure to do so is simple laziness.

2. Loss of computer battery: Similarly, running out of a computer battery during a dive would make it impossible to monitor your tank pressure. I have the same comment as above, however, which is to make sure your computer battery is fully charged before diving. I also use two AI computers, so if I were to lose batteries on one computer, I would still be able to monitor pressure using my other one.

3. O-ring failure: Just like with a high-pressure port plug, there is a static o-ring at the connection between the transmitter and the regulator first stage. This is extremely unlikely to fail provided that the transmitter is properly tightened onto the regulator. As with all o-rings, I would replace this one when I service the first stages and then forget about it. That said, I have blown this o-ring once. It happened when I pressurized the tank before diving, and it was because I hadn't actually tightened the transmitter into the first stage properly. The same could occur with any hose or port plug that is not properly tightened into the first stage.

4. The OPV: There is an overpressure valve on all transmitters, and it is possible to blow it. This would, again, occur when you are pressurizing the tank at the surface. I know of a few OPV failures that have occurred, and they all occurred dramatically and obviously when the diver was pressurizing his or her tanks before the dive. You are not getting in the water with a failed OPV.

5. Loss of comms: Older transmitters, as previously stated, did sometimes lose communications with their associated dive computers. This happened when other nearby transmitters interfered with the signal (either from another team member or from another sidemount tank), or, interestingly, when Bluetooth technology was being used nearby. Bluetooth speakers used at dive sites used to knock my comms out. So far, this is not an issue with the Shearwater SWIFTs. As the transmitters listen for other signals/transmitters and communicate when channel is clear, SWIFTs simply do not lose comms. Not a single person I know with SWIFTs (including myself) has even one incident of lost comms with SWIFTs.

Now, the interesting thing about transmitter failures is that with the exception of the two failures that would occur during gear set-up (and be dealt with then), the failures associated with the in-water use of transmitters are informational. In no way do these failures deplete the amount of gas in your tanks. Provided you have been running your gas management properly, you always have twice the amount of gas you need to get out of the cave at any point in the dive (usually more). While you can no longer monitor your pressure, this doesn't actually endanger you. Turn around and leave, and switch regulators periodically. The gas is still there. It's not leaking out. Compare this to the issues associated with pressure gauges failures.

I dive every day, and while I do attempt to maintain my gear properly, there's always going to be unexpected, statistically improbable failures if you dive often enough. I have had every single one of the failures listed below:

Photo courtesy of Natalie Gibb | Under the Jungle.

1. Pressure gauge needle becoming stuck: It's possible for an analog gauge's needle to become stuck. This happened to me once when my right tank was at 2400 psi. Luckily, I noticed that the needle was not dropping, and troubleshot the issue. I closed the tank valve and purged the regulator to depressurize it. The needle stayed at 2400 psi. I keep this gauge in my “gear failures” example box that I show to students. That gauge has been sitting at 2400 psi for about five years now. To me, this is one of the most dangerous failures I can imagine, because if a diver is not in tune with his or her gas consumption, it would be possible for the diver to use more gas than planned and endanger the life of the diver using the gear or his or her buddies in an emergency. At least a transmitter tells me when there is a problem.

Photo courtesy of Natalie Gibb | Under the Jungle.

2. High pressure hose failure: With enough use, any high pressure hose will fail. They usually do so with warning, and careful divers will observe a bubble or distortion under the rubber hose sheath and change out the hose. However, I have had a high pressure hose explode underwater with no warning. At this point your gas supply is slowly but actively being depleted, which is not a good feeling.

Photo courtesy of Natalie Gibb | Under the Jungle.

3. O-ring failure: Pressure gauges have three o-rings: the o-ring between the hose and the first stage, and the two tiny high pressure spool o-rings between the hose and the gauge. These all have failed on me, underwater, with no warning. In my opinion, the fact that the high pressure spool o-rings are on a swivel and move makes them more likely to fail than a static o-ring. Similarly, the pressure placed on the high-pressure hose as the diver manipulates it during a dive will invariably affect the hose itself and the connection o-ring between the hose and the regulator first stage (if the hose is not tightened into the first stage enough). In any of these scenarios, gas is now leaking from your sidemount tank and depleting your gas supply.

4. Shattered glass/plastic: The face of SPGs is exposed to impact, and I have had glass gauges shatter (during a dive, again with no warning). I have had plastic gauges crack. In some cases the failure caused a leak, and in some cases the safety gas held for the exit.

Photo courtesy of Natalie Gibb | Under the Jungle.

5. Inaccurate gauges: SPGs, with time, appear to be prone to inaccuracy. The calibration goes off and the gauge reads incorrectly. As I check all tank pressures with an independent pressure gauge before attaching my regulators, I have generally been able to ascertain when an analog gauge creeps out of calibration, but a diver would have to be pretty careful with his or her set-up procedures to catch this sort of issue. While it is theoretically possible for a transmitter to have the same issue, I have not experienced a calibration error in an electronic pressure gauge yet, so it is at minimum less likely to occur than with an analog gauge.

Clearly, there are potential failures with any method of monitoring your tank pressure. For me, the difference is how your chosen method fails, and what impact this has on your gas supply. I would rather be exiting a cave certain that my gas supply is not leaking out of my tank (even if I cannot read it), and confident in the knowledge that I have enough gas to exit, as opposed to being able to monitor my dwindling gas supply as it slowly leaks out of my SPG assembly.

Photo courtesy of Natalie Gibb | Under the Jungle.

Now, this article is about sidemount multi-stage cave diving, so consider that with every addition of a tank and regulator, a diver is increasing potential failure points. With four tanks and four transmitters, I have four potential battery failures/comms failures, and two potential computer failures, none of which affect the actual amount of gas in my tanks. That's six failure points. With SPGs, I have three o-rings, the hose, the SPG needle, and the SPG glass/plastic on each gauge. That's six failure points per an SPG, or 24 failure points, 20 of which will diminish my gas supply if they fail. To me, at least, I feel that I am both reducing failure points and reducing the gravity of a failure on multi-stage sidemount dives by using transmitters and no SPGs at all.

Benefits of AI for Sidemount in General, and Multistage Diving in Specific

Photo courtesy of Natalie Gibb | Under the Jungle.

Sidemount divers know that it's important to constantly monitor pressure in both sidemount tanks, and to switch regulators frequently to keep the tanks balanced and maximize gas in an emergency. With practice, gauge checking, and regulator switching, this becomes intuitive and second nature and it doesn't necessarily feel like gauge checking is a chore.

However, checking gauges does take time, and in certain scenarios, such as in long, tight restrictions, it can be difficult, or even impossible, to access a pressure gauge to check your pressure. Similarly, in reduced visibility scenarios (which are of course more common in sidemount as divers can physically access tighter caves that bring them in closer contact with the environment), it can be difficult to read analog gauges. For those with poor eyesight or restricted shoulder movement, reading an analog gauge can be difficult, and reaching the gauge to move it to a visible location can be physically painful.

AI and transmitters solve all of these problems, but that's not my main argument for them. In cave diving, divers constantly check their computers for depth, compass heading, dive time, and decompression status. Most cave divers hold their arms gently extended in front of their bodies, and a computer becomes like a dashboard in a car. The diver has it in his or her peripheral vision at all times, and is always aware of these vital dive statistics. With analog gauges, a diver has to choose (or remember) to look at the gauge. With the tank pressures displayed on the computer at all times, the diver is consistently and even subconsciously aware of his or her tank pressures at all times, with no additional movement or deliberate choice required. It's a big deal when running a reel, driving a DPV, or just navigating on a normal cave dive.

I would liken transmitters to other technological advancements in recent years. Sending a letter in the post didn't seem like a chore until email was possible. Sorting through cassette tapes and inserting one into your car stereo didn't seem like an issue, until drivers could create custom playlists on their phones and wirelessly link to their stereos. Sure, the post still functions and tape cassettes theoretically still work (if you can find one), but no one actually uses these antiquated methods because technology has made life much easier.

Transmitters are the same. You didn't realize you were suffering by looking at analog SPGs until you try transmitters. Then, it's nearly impossible to go back to the old technology. It's easier, less task loading, and I would argue safer, as you are aware of your exact tank pressure at all points in the dive. One of my divers who tried out my transmitters said it best after her first dive using AI, “this is necessary,” she said.

Taking this to stage and multi-stage diving, a diver now has four pressure gauges that he or she should be constantly monitoring because, even if a diver is not currently breathing from a tank, it's possible that a leak or free-flow could occur, and a quiet leak on a tank might not be noticed immediately. As most divers keep stage bottles that they are not currently breathing from closed, a leak could depressurize a stage regulator and allow water to enter the first stage if it loosens from the tank (which is more likely if the regulator is depressurized).

This task loading of monitoring multiple SPGs becomes more apparent the more tanks you add and the more difficult your dive plan is. Activities like scooter diving, exploration, and photography all increase a diver's task loading and therefore, the risk of missing a regulator swap or checking your tank pressures infrequently. AI solves these potential issues as the diver can monitor all the tank pressures at once by simply glancing at the dive computer.

Adding Transmitters to Stages When I Already Used AI

I have been an advocate of AI for several years now, and I happily used Shearwaters yellow and gray transmitter combination on my sidemount tanks. When I received the new SWIFT transmitters (I got two to compliment my yellow and gray transmitters) I was curious about how this would all work out.

My first concern was whether the yellow and gray transmitters would interfere with the SWIFTs, and whether once I upgraded my computer firmware for the SWIFT transmitters, if there would be any issues using the older models alongside the new ones. The firmware upgrade worked flawlessly. I was able to simply add the codes for the SWIFTs into my computer for AI #3 and AI #4 and immediately monitor both the new and old transmitters. There was zero issues, everything was backwards compatible, and it was fast and easy.

My next concern was how I would set up my home screen on my Teric and Perdix to keep track of four pressures. The solution I came up with was to configure my home screen pressures so that their physical location on my body matches the location on the home screen. I have my stages clipped below my sidemount tanks, so my home screen has the left sidemount tank on the left side of the screen, with the left stage below it. The right sidemount tank is on the right side, with the right stage number below that. I found this intuitive and easy to track. The mini-displays on the Perdix allow three pieces of information, so below the pressures I have my max depth on the left, and my average depth on the right. All the information is easily accessible and fast to understand at a glance.

My final major concern was stage drops. For those unfamiliar with stage cave diving, the cave diver plans to use a predetermined amount of gas from the stage bottle, then remove it, clip it to the cave line, and leave it there while he or she swims further into the cave. This reduces drag and profile, and makes it easier for the diver to access sidemount only passages that would not allow a diver with a stage to pass.

The concern I had here was what would happen when I swam away from my stage. Clearly, I would lose comms, but would comms reestablish when I swam back to the stage and retrieved it on my exit? So far, the answer is yes. As I approach my stage comms reestablish flawlessly. My only annoyance here is that while I am away from my stage, the “no comms” warning flashes in red in the stage tank pressure display and I find it distracting. It would be great to be able to confirm the message and have it turn white/ neutralize it. In the end though, I am picky and it's not really that big of a deal. It still infinitely better than having to deal with pressure gauges.

Photo courtesy of Natalie Gibb | Under the Jungle.

Rebreathers and Transmitters

While this blog is supposed to be about multi-stage diving in sidemount, I feel like it wouldn't be complete without discussing my favorite new use for four transmitters: cave diving with my JJ Rebreather and two sidemounted bailouts.

Anyone who has used a JJ knows that while the rebreather itself is amazing, the pressure gauge system for monitoring on-board gas is just miserable. Through necessity, the gauges are tucked away next to the diver's chest on both sides, and for a small person, getting them out to check the pressures in your oxygen and diluent bottles is a true pain once the gauges are tucked away behind the manual adds, bailouts, wing inflator, etc. I'm four-foot eleven inches tall, and I just don't have that much space on my chest.

I heard of people who simply get rid of their O2 pressure gauges (the horror) or those who don't regularly check their onboard gas pressure, but this in not an option in caves that require nearly constant ascents and descents, which necessitates that the diver constantly vent and add gas to his or her breathing loop. I did about two dives on the JJ before I gave up and put transmitters on the on-board gas.

Photo courtesy of Natalie Gibb | Under the Jungle.

However, there was still the bailouts to monitor, and although they should be at a constant pressure through out the dive, a leak could go undetected if a diver doesn't periodically monitor bailout pressure. I put my other two transmitters on my bailouts, and now I can easily monitor all four tank pressures with no stress or task loading. It doesn't get much better than this.

Teaching Cave Diving

Finally, I have found benefits to teaching cave diving in sidemount with transmitters. The first, sneakiest one, is that I can read my student's pressures by furtively glancing at their computers without them knowing. This allows me to assess whether they are balancing their tanks properly and sticking to their gas plan. It's great, because I don't have to disturb the student by physically manipulating their gauges or swimming over to them during the dive.

In cave diving instruction, teachers will often play “the ghost.” We hover darkly above the team of students, attempting not to disturb them as they run their dive, stepping in only when behaviors or procedures need to be corrected. We keep our primary lights off, and if we are playing the ghost well, students are not disturbed by our presence. With analog gauges, I would have to carefully turn on my primary light to see my own gauges, and this would invariably cause light movement that would distract the student. With AI, I can just glance and my computer and be a much more effective ghost!

Photo courtesy of Natalie Gibb | Under the Jungle.

Big Dives with AI

I have experienced the most benefits from AI in my most extreme dives. Exploration, multistage DPV diving, and extreme sidemount caves really do improve when you don't have to stop everything you're doing and look at a pressure gauge.

When I need to be 100% focused on the environment to be safe and effective, AI shines. My favorite configuration for big, reduced visibility, end of line exploration pushes is to use sidemount tanks configured normally, with a left stage regulator looped from left to right over the back on my neck, and a right stage regulator with a reversed second stage looped from right to left over my neck. I scooter to end of line (where I will start exploration) treating the stage tanks like an additional set of sidemount tanks. I monitor the tank pressure using transmitters, and switch back and forth between the tanks without going off the scooter trigger. This keeps the tanks balanced and reduces the sideways list caused by disparate pressures in the stages, and I can be sure there are no issues with freeflows from my sidemount tanks as I zip at high speed to my end of line. I drop both scooters and both stages when I reach my exploration start and then off I go into the unknown.

Photo courtesy of Natalie Gibb | Under the Jungle.

To AI or Not to AI?

After my experiences with transmitters, I can say that I have high confidence in the technology, and trust transmitters more than I would trust traditional gauges. I reduce my failure points by only diving transmitters, and manage my gas (as all cave divers do) so that at any point during the dive I have at least twice the amount of gas required to safely exit. If I ever was to lose comms, I would simply switch regulators every few minutes, and head safely out, confident that I am not losing gas from a physical failure. Task loading, specifically with monitoring the pressures of more than two tanks, is almost negligible, and I believe that AI is the future of diving.

The only drawback of transmitters is the price, but scuba diving, and specifically technical diving, is an expensive sport. With the risk inherent in the activity, it makes sense to purchase the safest, most robust, and easiest to use equipment that money can buy. The benefit of transmitters and AI is directly proportional to the complexity of and extremity of your diving.

8 Responses

  1. Jozef Koppelman

    Just made the switch. Thanks for a very thoughtful, well written discussion.

  2. Thanks, Jozef! It was fun to write and came from the heart 🙂

  3. Jason N

    Loved your article and agree 100%. I am still saving up for mine!

  4. Pete Tanner

    Great article, and note your comment ” I also use two AI computers, so if I were to lose batteries on one computer, I would still be able to monitor pressure using my other one.” So if using a Perdix AI and a Teric with black and yellow transmitters, can both be connected to both computers at the same time, or would you have to be connected to your main, then enable AI on the backup if the primary failed?

  5. Dahab Boris

    Nice article. Re the JJ just route the SPG’s over your shoulders (much better than by your bottom) Easy to glance at if you don’t want to use AI

  6. Hi Natalie,

    I have been thinking about the switch for the past two years, but I was hesitant to do it for the same reasons described in your article and the high price of the transmitters. A few of my friends use the transmitters and love them. After reading the article about your own experience, I decided to go for the transmitters.

    Thanks for sharing your experience.

  7. laurent

    Would love to have an answer about multiple connexion of 1 transmitter to 2 computer

    • Shearwater Research

      Hi, you can pair 1 transmitter with two computers. You can enter that transmitter as T1 on both computers.


Leave a comment