Probably you have many questions. I know this because I’ve heard them all, in the course of this project. For example, “How do they get air?” Via an air compressor on the surface sending fresh air down constantly. “What if it fails?” The backup kicks in. “What if there’s a power outage?” Both pumps have independent battery backup.
Then there’s “how do you feed them?” I can either drain the water and open the lids, or decouple them from each other after closing the water tight gates by hand (using hard to see handles in back) to prevent water entering, then bring the modules to the surface. This is also how I clean them out periodically and replenish the water.
Yes, I know it’s cold underwater, the floors are electrically heated. Yes, I know the air will be humid, there are in-line moisture traps to remove humidity from air before it goes down to the habitat modules. I know it’s working because there’s a combination thermometer-hygrometer inside that monitors temperature and humidity.
What’s that? I should add a moisture sensor to each habitat which sounds an alarm if water enters? What a good idea! Such a good idea in fact that I already thought of, and implemented it, many years ago. The idea for this project first occurred to me in 2010, and it took me many months of research before I felt confident I could build anything safe enough to put a living creature into. Consequently, none have drowned.
Can the habitat modules implode? No, they’re what’s known as “positive pressure”, like a diving bell. Not “one atmosphere” like a submarine. Diving bells don’t structurally resist water pressure. They maintain air pressure inside equivalent to or higher than outside water pressure. This is how they can have open hatches in the bottom to enter or exit through without water rising to fill the interior.
Submarines instead maintain one atmosphere of pressure (equivalent to sea level air pressure) inside at all times, using the structural strength of the hull to resist increasing water pressure as depth increases. That’s a very precise, exacting engineering challenge which is why I didn’t do it that way. Positive pressure habitats, conversely, are very easy to build safely, and inherently fault tolerant.
What I mean by inherently fault tolerant is that the internal atmospheric overpressure does much of the work of preventing water ingress for you. See those bubbles? Many panic at first glance, reasoning “if air can escape, water can enter” which sounds perfectly reasonable despite being totally false.
Water can’t get in through the seal (where the bubbles are escaping through) because it’d have to get past the higher pressure escaping air. As a thought experiment, imagine carrying an inflated balloon underwater. If you allow some air to escape, the balloon decreases in size, but does water enter?
This principle is what makes positive pressure habitats so inherently safe. There’s been about 70 human scale undersea habitats in history, since the mid 1960s. Nobody has ever drowned in one. The only time water entered one, it’s because it tilted severely while being lowered to the seabed (Sealab I). Contrary to that one episode of LOST where Charlie drowns in the Looking Glass habitat, even if a window broke, water would rise only to the top of the window. There’d still be air trapped above that to breathe from.
Having satisfied you that I am not a maniac, that I take extreme precautions for the safety of my beloved pets and that with correct engineering it is possible to support mammalian life underwater in comfort and safety, we can now discuss how the project began and why I undertook it to begin with.
As you can see in the photo above, I started out using off the shelf products with a very naive idea of how easy it would be. That’s a habitat intended for hermit crabs. I scrapped it almost immediately. Likewise I underestimated how much solar and battery capacity would be needed to do this outdoors.
I began the project because it’s an idea I had as a child watching Rescue Rangers, the Secret of Nimh and from “Down the Tubes”, a level from the original Earthworm Jim for Super Nintendo and Sega Genesis set in (surprise) an underwater hamster colony.
There’s also this experiment performed in 1964 wherein a habitat was submerged with a single wall made of microporous silicone to act as an artificial gill, but I didn’t find out about it until the project was well underway:
Mostly it was just a fun idea and seemed feasible from an engineering standpoint. I was also very depressed at that stage of life though and simply needed a hobby.
Hampture gave me something to do. Something that would stimulate my mind. A project that I could directly see my labor resulting in progress on, everything coming together a little bit more, day after day.
Before that, each new day seemed identical to the one before it. Projects disrupt this cycle by giving you something by which to identify the passage of time.
I bought a bunch of books about marine engineering and the history of manned undersea habitats, to learn about the engineering principles involved. Pretty soon I had a small one room prototype. Obviously not intended for long duration use, just a proof of concept and test bed.
This prototype was adjustable to either be one atmosphere (hence the air return tube) or positive pressure (accomplished simply by turning a valve in the air return tube to block air so it would need to escape through the habitat lid’s seal). I was unsure at that time which approach I would ultimately commit to and wanted the flexibility to try both.
The compressor was a deep water certified aquarium pump, rated to push air down to a depth of 8 feet. This depth has only been approached twice in the history of the program. The first time was when I tested the 1 atm mode at the bottom of a local public outdoor swimming pool.
After a successful test with no occupant, I proceeded to repeat the test, this time with the hamster inside. I was stopped and questioned during this process by a security guard, but he was so confused, maybe amused, by my explanation that he did nothing to stop me:
The power supply was a portable battery of the type commonly used to jumpstart cars. It could run the pump for about five hours before it started to struggle, as established in unmanned (unhammed?) tests. All habitats, throughout the history of this program, were extensively tested without animals inside first.
The pool test didn’t really contribute any data in this regard since it was very brief (as I wasn’t supposed to be there at that hour, much less submerging a hamster) but it did confirm the operational depth of the air compressor.
1 atm mode worked only to a point. If the pressure needed to cycle stagnant air back up the return tube exceeded outside water pressure at depth, it simply bubbled out through the seal. That’s when I decided to abandon the 1 atm approach moving forward.
There’s the little fella inside the contraption. I named him Megafucker Supreme, because it seemed like a history making pioneer should have a suitably epic name. It’s hard to really call this thing a habitat. More of an improvised diving chamber.
It nevertheless had food and water dishes, litter to absorb waste, an LED light, a battery operated resistive heating pad, and a camera. Having tested it extensively in the tub without a hamster inside (then a few times with him in there) eventually I felt confident enough in the design to submerge it in the local man-made lake.
This would be the first test of the habitat (or diving chamber, if you prefer) in a natural body of water. I had many concerns but most involved other people’s curiosity, handled by being physically present the entire time to answer questions and prevent equipment theft:
He seemed much calmer this time, which I attribute to the ambient lighting. It was dark during the pool test. Although hamsters have poor eyesight, ambient light tells them something about their surroundings, information that’s absent in the darkness. The blue LED furnished just enough illumination to film by.
The lake test taught me a few things. Mainly that most bodies of fresh water I might choose to submerge habitats in would have such poor visibility as to make the endeavor nearly pointless as nothing would be visible outside. If I lived by a tropical lagoon, it’d be a different story, but that wasn’t in the budget.
The next step, obviously, was to greatly expand the habitat and to perform another outdoor test. I quickly got to work building a four room habitat where the rooms were all linked by clear acrylic tunnels. They had to be permanently mounted onto the same platform as, at this time, I’d not yet worked out a modular tunnel system that could be connected or disconnected as desired.
There’s the original prototype in the lower right for size comparison. Every room got its own dedicated air inlet to prevent the formation of CO2 pockets. There was a webcam inside this time which I could hook up to a laptop on the surface to monitor the main chamber interior in real time.
There was a secondary keychain camera with terrible quality I got for a few bucks off Amazon to stick in one of the 3 smaller modules, just so I could get some shots of hamsters moving from one module to the other. Funding was tight in those days, concessions had to be made.
The top few feet were unseasonably good vis, but the muck collected near the bottom, making outside visibility in this pond about as poor as it had been in the lake. The new 4 module habitat prototype nevertheless performed as expected. I left it overnight and collected it the next morning, this time using an amply oversized battery pack. It was a nontrivial risk, but the location was remote enough that I wasn’t too worried.
What I learned from this test was that if I wanted to go any bigger, I needed a modular tunnel solution. Mounting all the modules permanently to a single platform so the tunnels didn’t flex wasn’t a long term solution, I needed to be able to connect and disconnect modules while submerged without allowing water inside.
This is around when I began to experiment with other habitat designs inside a big transparent tub of water, not having sprung for a proper aquarium yet. There were many configurations I tried out during this period, knowing I wanted more than one module and a clear tunnel, but not yet certain how to implement those features.
None of these experimental habitats were in use for long periods. Hamsters need a lot more than just warmth, fresh air, water and food. They need an exercise wheel, they love to run run run. They need space, which they get plenty of in their enormous, conventional land cage where they spend their time while not underwater.
They also need wooden chews, and multiple hiding spots. They need nesting fluff to burrow in. Without those things they would quickly accumulate nervous energy with no outlet, and their teeth would become dangerously overgrown.
Taking all this into account, even as I built smaller test habitats to see how specific design features fared in a positive pressure underwater environment, I gradually moved towards conditions that would meet a hamster’s needs. Not just for survival and comfort but also fulfillment.
This meant the ultimate form of the habitat would eventually need nesting fluff, an exercise wheel, wooden chews, at least two hides, heating, dehumidification, a water bottle, food dispensers, a modular waterlock system for connecting them together, and more. This brought an end to the experimental one-off habitat era of development:
…and refocused the project, going forward, towards the development of a habitat that would combine all the features of the prior prototypes. A tall order! But who better to make it happen than myself? I don’t know of any other comparably qualified underwater hamsterologists in the world.
This refocusing prevented resource waste, as I now had a more or less clear picture of what I meant to build, so I wasn’t building tangential prototypes any longer. It also meant I could move towards longer duration submersions, once all the comfort and happiness boxes were ticked.
The single biggest problem that needed solving was how to connect multiple habitats together in a water tight matter such that I could connect and disconnect them while submerged, without letting water get inside. I ultimately settled on a system where the tube is permitted to flood, purged with compressed air once screwed into place.
Tub tests began with no animals, then progressed to involving animals once proven, so I could see how they liked moving through the clear tunnels. I shortened it considerably in the end, for lack of space in the aquarium I’m still using today. The black rubber band covers the holes on the top and bottom where air is pumped in, and displaced water escapes, during purging.
At last, with this hurdle cleared, there was theoretically no limit to how large a submerged habitat complex could be. That is, if I had a nearby body of water to put these habitat modules into. Without one, I’ve been limited to just the space available in the aquarium, which comfortably fits only two habitat modules linked by tunnel.
That more or less is how the project went from a single cramped testbed chamber to the dual module fully apportioned habitat complex it is today, with heated floors, dehumidifiers, nesting fluff, a wheel, moisture sensors/alarms, redundant air pumps, battery backup and all the other design features I added along the way.
The Bioshock theme is just for funsies. The gravity fed food dispensers are shaped like vending machines, I 3D printed them. Those black cases on the bottom are water tight ballast pods which prevent the lead weights inside (20lbs each!) from contaminating the surrounding water. Hard to believe everything it took, in the end, to make real what I thought would be an easy, fun short term project.
Did I get everything out of it I hoped to? Yes, and more. Besides helping alleviate my depression, a funny thing happened along the way. People took notice of the blog I started to document this process (now abandoned for Patreon). Money started coming in as people who had the same idea as kids, but never the lack of common sense needed to seriously pursue it, wanted to support the internet madman making their childhood dream a reality.
It wasn’t much for a long time. Not even enough to pay for the components. But then a controversial internet personality, Mr. Metokur, took notice and gave this project a shoutout to his enormous fanbase. Funding skyrocketed virtually overnight:
This utterly and abruptly changed my life. With Patreon income added to the ongoing (at that time) Steemit income I could now afford to stop driving for Postmates to make rent. I could upgrade my computer, buy proper cameras instead of the ancient potato tier ones I’d been using until then and more generally fix a lot of stuff I needed for the project which had been broken for a long time. But really, my life was broken too and now I could at last begin to fix it properly.
I’d already bought a car, having profited handsomely from Steemit in its heyday between 2016 and 2018 but now I could afford more than basic liability insurance. I could afford health insurance, at last getting my respiratory problems looked at. It included dental insurance, which I recently came to need as I fractured a tooth and will need a root canal.
After losing my cat Jenny of 15 years back in November, followers helped pay the ruinous end of life expenses, cluing me into the wisdom of insuring my next cat, All Ball. So I have pet insurance now. My equipment is insured as well, I can afford a gym membership, I can eat better, and in all other ways my life has improved.
I worked hard for a long time but was also very lucky. I’d be remiss not to admit the huge role that luck played in all of this. It couldn’t have come at a better time either. As an autistic adult I’m part of a demographic that has a hard time getting and holding down conventional 9 to 5 jobs. Depending which study you trust (the commonly cited 85% figure is from Autism Speaks, ick) around 70–80% of us remain unemployed longterm.
I had a long string of shit jobs prior to this successful project, most of which I lasted less than a year at before they fired me. Two were call centers, a bad fit for an antisocial person but I did a good job according to post-call customer satisfaction surveys, I just had to work around the company provided flowchart in order to achieve that.
This sort of mismatch (or miscommunication) was typical of my sordid work history. I could do a good job, or work within the system as they designed it for us, but not both. All of that’s behind me now, however. I spun this one success off into several others including merchandise, a Youtube channel, a Twitch stream, even an upcoming children’s book and computer game.
But even as the project attracted positive attention, it also attracted negative attention. It came under fire from a diverse range of internet subcultures such as furries, vegans, and hamster enthusiasts who naturally took one look at cute hamsters in seemingly dangerous underwater enclosures and assumed the worst.
Having learned to support mammals comfortably and safely underwater, something my autism arguably makes me well suited for, I would now have to learn how to manage large nebulous internet mobs of angry mammals, something I’d argue that same autism made me especially terrible at.
Even within the subset of negative reactions I’ve gotten, the range of different takes and especially reactions to disclosure of my neurodivergence have been very interesting, and telling. But that’s a matter for a different article. I’m satisfied I’ve brought you up to speed with the project, such as it is now. Stay tuned for further articles on this topic, if there seems to be sufficient interest.
All the links to my Twitch, Youtube, merch and everything else can be found right here. Consider supporting me on my Patreon if this is the kind of creative weirdness you’d like to see more of in the world.