My Bachelor Thesis is based on creating a self-regulating Aquarium. Can you guys help me brainstorm a plan?

[Paul B]

The problem with that approach might be that it would be a bit too easy i would think. "I bought an auto feeder, a light, a jar and some inverts, put them together and nothing died" doesn't make for a very long or impressive thesis. Talking page after page about fine tuning amount of food and light by hand to the extent that you can see after 2, maybe 3 months might stray away too far from mechanical engineering and "regulating".

You are probably correct. I didn't have to go to college (I was smart enough already)
So I never wrote a thesis.
But we still need more information as to exactly how complicated this is supposed to be.
I wrote a book, it is 240 pages, he could copy it word for word. I won't tell.

 

[58e970b2-3f88-4897-87ba-5]

@58e970b2-3f88-4897-87ba-5 1st: love username and avatar.
Do you pump up from bottom to top? How do you control the siphon/drain?
Are you using plastic on bottom tank to reduce evaporation?
Maybe something like a Clarisea instead of socks. Magnetic stirrer for detritus?
Bubbler for gas exchange?
When you say "flow is controlled by coral position" do you mean you observe the coral position and you adjust direction and velocity of bubbler or other pump outlet?
-This is a fantastic example-
Can you list what params you are concerned about for water quality?
Salinity, Ca, Mg, alkalinity, pH, trace elements, nitrate, phosphate, organic and inorganic pollutants?

-> Do you pump up from bottom to top? How do you control the siphon/drain?
Pumping form bottom up to the top. Here is a better view of the internal plumbing.

The top jar is a drink dispenser (came with a 1/2 inch hole in the side). I have a pipe going in there and up towards the surface for the drain (it can be a bit noisy). I have the hose from the pump zip tied to the drain pipe.
When the pump is out, the water siphons back though there (pipe is just high enough to keep to Candy Cane submerged)
Bottom jar was sized to accommodate all the water that siphons out.
My main worry is if the drain clogs. Unfortunately there is no backup.
I forgot to say that part of the maintenance is cleaning algae out of the drain.

-> Are you using plastic on bottom tank to reduce evaporation?
Bottom tank has plastic wrap over it to reduce evaporation. Also need to keep a dip in the center to reduce condensation from dripping out.
Main issue is I have gaps around the lid and bottom plastic. Lose about a cup of water each week due to evaporation. If I added an auto top off or sealed this better I could keep things happier and more stable.

-> Maybe something like a Clarisea instead of socks. Magnetic stirrer for detritus?
Good ideas. I find detritus normally accumulates in the back corner opposite to the pumps (hard to reach spot).

-> Bubbler for gas exchange?
Yep. I originally started without the bottom sump and just had the bubbler for flow and gas exchange, but ran into some issues with occasionally overfeeding and added the sump so I could run a filter sock easily.

-> When you say "flow is controlled by coral position" do you mean you observe the coral position and you adjust direction and velocity of bubbler or other pump outlet?
I meant that I observe the flow and place the coral where I think they will be happy. With the current setup I do not have a lot of control over flow.

The pump is a basic one with only on/off setting (a different pump could give this option if needed).
To control flow a bit I can adjust the valve on the bubbler and reposition the head. I have it currently mostly under the pump outlet because it provides a bit more randomness to the flow.

-This is a fantastic example-
Can you list what params you are concerned about for water quality?
I monitor salinity (and trust my salt mix is mixing to what it states it does). Basically check that the water for the water change is the right salinity, then I have a line drawn on the side of the bottom tank to help track and replace evaporation.
I have checked nitrate a couple of times recently (got a bigger than that requires I test regularly). Nitrate is consistently at 0, so I may start working on increasing this to see if I an increase growth.

 

[elysics]

You are probably correct. I didn't have to go to college (I was smart enough already)
So I never wrote a thesis.
But we still need more information as to exactly how complicated this is supposed to be.
I wrote a book, it is 240 pages, he could copy it word for word. I won't tell.

Oh no, wasn't downplaying you at all. Your accomplishments are awesome

Just saying that the thesis is probably supposed to be an engineer's take on this, not a biologist's or chemist's. In particular about regulation, i.e. measuring values and then correcting them, in a loop. And focusing on the overall system rather than particulars of any individual part of it would be either trying out something completely new, which 4 months is probably too short for, or redoing what others have already done, sticking a PID on everything, building a controller from scratch, etc.

 

[58e970b2-3f88-4897-87ba-5]

Not sure if this is related or useful, but you could touch on some issues related to equipment malfunctions and testing errors.

An example would be automated water top off. If I add water based on a level sensor, the sensor failing could kill my tank and flood my house. If it knew that I regularly add about X ml of water per day, it could detect a large variation from this an stop before causing major issues.

If you are dosing something like calcium hydroxide (kalkwasser) to maintain calcium level and using an automated method to detect how much is needed, then you could watch for sudden changes in usage. This might indicate a testing problem (or a sudden decrease in usage could mean your coral have stopped growing due to another issue).

 

[Paul B]

Oh no, wasn't downplaying you at all. Your accomplishments are awesome

I was just messing with you. You could downplay me all you want there is an entire plethora of things I have no clue about.

 

[KrisReef]

@Paul B is correct.

Does the saltwater aquarium in this Engineering thesis have to keep anything alive, or just be "automated"?

My system for water temperature control has a heater and a chiller. An engineer might want to use a heat pump.

Water "filtration" would be simplest (and automated) with a rollermat (for as long as the roll should last).

Bio-filtration needs to match the biological loading in the system, if any? Rock or bio-balls have long been proven to work well for this application.

Lighting probably should be provided unless this is perhaps a Mariana Trench habitat? A light source and timer are nice engineering focal points.

Continous water changes, automated saltwater mixing and water change replacement(s) can provide for an almost perfect design to replicate natural water flow and constant water exchange on a living reef.

Keeping all this water in motion will tend to keep oxygen levels high, and that may also be where your grade level attains, but we need to know if this tank will have to be constructed and demonstrated, or if this assignment purely a text design exercise?

 

[jta117]

What do you mean by self regulating?

Is the end goal an aquarium that you never touch? Just a tank with a bunch of machines doing all the “work” on it?

Yeah, pretty much a Smart Aquarium. Of course, never touch is a big ask and probably not practically possible but the target is to get as close as possible to that.

 

[jta117]

Correction: make it simpler than you think you should and in the unlikely event you have extra time you can add to it.
It's all about the water...and light for corals, algae, anemones, and other photosynthetic organisms.
Dont forget about oxygen and CO2. If you have lots of flow, only needed for coral or some types of filtration, gas exchange happens, but if low flow, you will need to do something about it.
Remember, you're not going to be graded on how well you understand aquaria but on how well you apply what you've been taught. Focus on the ME aspects: DiffEq, controls, thermo, heat transfer, maybe fluid mechanics (pipe flow and head loss), instrumentation, power, etc. Chemistry is important but it's not your area. Don't get bogged down by it. Chemistry is a great place to simplify.
Keep it simple.

I had a meeting yesterday with my Professor and to be honest, It seems he wants me to gather as much theory as possible and not focus on the setting up of the actual aquarium which bummed me out a bit. That part will be done by another person.

 

[jta117]

There is a lot to know about aquariums, a fish only would simplify your thesis but it still is quite complex. Anywhere from parameters, filtration, and care. Without proper setup fish tanks could not possibly survive in our homes, everything from having to establish bacteria and biodiversity to the types of food we feed. There is lots to know, just ask and most people here are very experienced and could give you a definitive answer

Yep, Im starting to notice the vast amount of information i would have to gather here. And honestly, all the people here have been immensely helpful and Im trying not to take advantage of that with asking a question everytime i hit a bump.

 

[jta117]

How long does this tank have to last? Is it only for a few months? If it is, that is easy and you won't need to add anything except food. Can you get away with just fish, crabs, shrimp etc. and no corals?

If so, it is immensely more easy and as I said, you won't have to add anything but food. You can also use any light you like as fish can get by on light that you can see by and nothing special like spectrum, par, lumen s etc.

The thing is, I wasnt very clear with the original post. I need to design a system that would be adaptable to all kinds of things. So basically you just input the tank size, the types of fish you have and it would probably do the rest. This is turning out to be much much much more difficult than i first thought. How the hell am i supposed to research all that information in 4-5 months? I mean there are millions of species (probably, idk) and i have to create profiles for groups of fish species or corals.

 

[jta117]

@Paul B is correct.

Does the saltwater aquarium in this Engineering thesis have to keep anything alive, or just be "automated"?

My system for water temperature control has a heater and a chiller. An engineer might want to use a heat pump.

Water "filtration" would be simplest (and automated) with a rollermat (for as long as the roll should last).

Bio-filtration needs to match the biological loading in the system, if any? Rock or bio-balls have long been proven to work well for this application.

Lighting probably should be provided unless this is perhaps a Mariana Trench habitat? A light source and timer are nice engineering focal points.

Continous water changes, automated saltwater mixing and water change replacement(s) can provide for an almost perfect design to replicate natural water flow and constant water exchange on a living reef.

Keeping all this water in motion will tend to keep oxygen levels high, and that may also be where your grade level attains, but we need to know if this tank will have to be constructed and demonstrated, or if this assignment purely a text design exercise?

Thanks for your reply! Umm it would definitely have to keep something alive. It will basically be a basis thesis for an electrical engineer who will use it as a guide for building and connecting the physical parts. Yes, it is theoretical but will be applied in the future by someone else.

" if this assignment purely a text design exercise?" yep... which kinda bummed me out. I would only be responsible for gathering all the theory necessary for the software to automate everything.

 

[xiaoxiy]

This is turning out to be much much much more difficult than i first thought. How the hell am i supposed to research all that information in 4-5 months? I mean there are millions of species (probably, idk) and i have to create profiles for groups of fish species or corals.

Without a doubt, this will be a pretty hard endeavor, but you won't have to create separate profiles for each different livestock. A lot of them can be grouped together as they have similar requirements.

I would figure out what size of aquarium you wish to automate, as this will dictate livestock choice and can help dramatically limit the amount of livestock you need to read up on. Most corals can be generalized under the broad category of soft corals, LPS corals and SPS corals. For your first reef (especially as it's your first attempt at automating things), I recommend soft corals and LPS. SPS can be finnicky, especially in a newer tank.

Most of the common reef parameters can also be generalized into broad ranges as well. The two parameters I would focus on monitoring/automating/regulating regularly would be alkalinity and temperature. If you're doing balanced dosing to replenish inorganics (alk, calcium, magnesium and trace), you will really only have to monitor alkalinity as calcium and magnesium dosing will be directly related to alkalinity consumption. Temperature is vital too. While salinity is important, it only has to be measured occasionally (typically when you mix up new salt water).

 

[jta117]

You are probably correct. I didn't have to go to college (I was smart enough already)
So I never wrote a thesis.
But we still need more information as to exactly how complicated this is supposed to be.
I wrote a book, it is 240 pages, he could copy it word for word. I won't tell.

I had a meeting with my professor yesterday where he cleared a lot up and also ticked me off.

The objective of the thesis is providing the theory required to build a model to automate an aquarium. So I need to research the important parameters that need to be controlled and research the co-dependency of these parameters on each other. Find out which fish require which parameter ranges and that kind of stuff. I dont really feel like an engineer right now :/

 

[jta117]

I would figure out what size of aquarium you wish to automate, as this will dictate livestock choice and can help dramatically limit the amount of livestock you need to read up on. Most corals can be generalized under the broad category of soft corals, LPS corals and SPS corals. For your first reef (especially as it's your first attempt at automating things), I recommend soft corals and LPS. SPS can be finnicky, especially in a newer tank.

Thats the thing. I dont have a fixed size of Aquaria. This is aimed for a wide range of customers. You just have to input your tank size in the app and all the inhabitants of the aquarium and based on that, the software would automate everything. This is purely theoretical at this point and dont know if it is possible but thats the aim of the thesis. To create a product that automates your aquarium.

 

[58e970b2-3f88-4897-87ba-5]

I had a meeting with my professor yesterday where he cleared a lot up and also ticked me off.

The objective of the thesis is providing the theory required to build a model to automate an aquarium. So I need to research the important parameters that need to be controlled and research the co-dependency of these parameters on each other. Find out which fish require which parameter ranges and that kind of stuff. I dont really feel like an engineer right now :/

...No one said you can’t setup an aquarium.
Just remember cost scales with size. Something like a reef jar or fluval evo are generally your best value options.
...also remember you still have a thesis to write. Don’t get to distracted by an aquarium or setup something too hard/complicated to maintain.

 

[keithIHS]

I hear your pain, and have experienced it, but if I may stray into the philosophy of engineering (and life) for a moment...your prof is your customer, and very often your desires and the desires of your customer won't match. OK, yes its galling that you are actually the one paying, but trust that this challenge really is beneficial. Also, frequently your customer's needs and your skills wont line up either, and it can be a dilemma deciding when to attempt to deliver and when to admit you're not up to the task. In this case you're being tested on your ability to get outside your comfort zone, perhaps intentionally. Remember that your prof may be an excellent reference in the not too distant future. And you want an A, or should.
One of the biggest challenges in engineering is simplifying the problem appropriately. You want sufficient accuracy within the constraints of time, money, and resources. This will require ignoring some things, probably lots of things. Make those assumptions, document them, and move on. Then brief your customer periodically (more often than they require can be a good thing, especially early in the project) and adjust course as needed.
Real world problems are far more complex than homework problems. Do what you can to simplify and break the problem down into homework-sized problems. Make a schedule! Give yourself some slack (margin, safety factor, empty time) in your schedule for the unexpected and your lack of skill in this area and in lack of skill in making schedules.
Maybe we should move this to PM? Haven't done that so not sure how but if you want and others want it gone from this thread (which is very reasonable) I can figure it out.

 

[KrisReef]

I hear your pain, and have experienced it, but if I may stray into the philosophy of engineering (and life) for a moment...your prof is your customer, and very often your desires and the desires of your customer won't match. OK, yes its galling that you are actually the one paying, but trust that this challenge really is beneficial. Also, frequently your customer's needs and your skills wont line up either, and it can be a dilemma deciding when to attempt to deliver and when to admit you're not up to the task. In this case you're being tested on your ability to get outside your comfort zone, perhaps intentionally. Remember that your prof may be an excellent reference in the not too distant future. And you want an A, or should.
One of the biggest challenges in engineering is simplifying the problem appropriately. You want sufficient accuracy within the constraints of time, money, and resources. This will require ignoring some things, probably lots of things. Make those assumptions, document them, and move on. Then brief your customer periodically (more often than they require can be a good thing, especially early in the project) and adjust course as needed.
Real world problems are far more complex than homework problems. Do what you can to simplify and break the problem down into homework-sized problems. Make a schedule! Give yourself some slack (margin, safety factor, empty time) in your schedule for the unexpected and your lack of skill in this area and in lack of skill in making schedules.
Maybe we should move this to PM? Haven't done that so not sure how but if you want and others want it gone from this thread (which is very reasonable) I can figure it out.

That was a very sincere and hopefully helpful response to the OP. Reef2Reef gold.

 

[bherbold8]

Hey guys, I really need your help. I am a mechanical engineer tasked with designing and configuring a self-regulating Saltwater Aquarium but unfortunately I know nothing about aquariums (Well, not true, ive been reading up a bit, but id still consider myself completely ignorant on the subject).

To clarify a bit more, right now I am still in the research phase and to be honest, I am not finding that much success, not that there isnt any information but for the fact, that I dont really know what exactly NEEDS to be regulated. Ive been brainstorming for almost an hour now after some basic reading about Saltwater aquariums and so far this is what Ive got.

Theory behind the Aquarium

• Parameters to control
• Water Quality
• Temperature of the room?
• Feeding
• Imbalances of the biological system
• Types of Saltwater Aquariums
• Water Filtration
• Biological Aspects behind the aquarium (e.g. Cycling a Saltwater Aquarium or Nitrogen Cycle)
• Sensitivity Analysis
• Size of the Aquarium and how that affects the parameters that need to be controlled.
• Light sources

For the first point (Parameters to control) Ive seen a couple of sites mention around 7 or 8 important parameters but those are all concerning the water. Is water really the only important thing that needs to be regulated in an Aquarium? The rest of the list, Ive been basically coming up with using common sense but I dont know how true they are. Can any aquarists here help me out with the brainstorming? A good question to prompt some good answers would be, for example, what do you do daily to your Aquarium? Literally anything you can think of, not just maintaining the water balance. and of course if youve got some other ideas, it would be immensely appreciated.

Oh man good luck hah you'd be competing against the saltwater tank industry but I'll follow along!

 

[Paul B]

The objective of the thesis is providing the theory required to build a model to automate an aquarium. So I need to research the important parameters that need to be controlled and research the co-dependency of these parameters on each other. Find out which fish require which parameter ranges and that kind of stuff. I dont really feel like an engineer right now :/

That makes it much easier. Not for me as I don't use any technology but many people on here automate everything so one of them should be able to help you. There are 5 or 6 chemical parameters that can automatically be checked and have the appropriate chemicals added. The temperature, water level, salinity etc, can also be automatically controlled as can water changes and lighting schedule as well as Par values.

You can really go nuts with this stuff so you should be able to find quite a bit of technology to write about.

 

[elysics]

Look up the principles behind devices like the ReefBot, mastertronic, kh keeper, trident. Throw in a combination of temperature controller and heater, maybe figure out some nifty way to automate salinity readings (those probes suck), a general ph probe, and a bunch of dosing pumps and automatic water changer. Maybe even include some way or protocol to regularly send in water samples to an ICP testing company or your own dedicated ICP machine.

If you are only concerned with the theory and not actually building (or paying for) any of this, that makes it much much easier.