CONTEXT: I'm working on a project consisting of making an underwater robot "fish" for an aquarium. It is mandatory to get at least 8 hours of autonomy. I'm responsible of the electrical part of the project and during all of my bachelor degree I've never got any lecture on batteries and I have no idea on how to get what is required. But I managed to do some research and try to figure it out:

  1. I took the power consumption of all the potential components necessary for the project (motors & pumps, Usound captors, ucontroller and a set of RF receiver and transmitters) from the specs and datasheets available.
  2. I got a battery of a 1200 mAh capacity. The battery is originally used for quadrocopters, i figured it would be good for the project since it's relatively the same thing (a software and motors).
  3. Thanks to some references (Battery university, Eletropaedia), I estimated the battery autonomy. I used an ideal estimation [=Battery Capacity/sum of power consumptions] and I supposed all the components worked 100% of the time.

The results I got is far bellow needed :(

Observations :

It seems that the pump is the only component to have an excessive consumption

Difficulties :

  1. Since I have no experience in the field, I don't know if the battery I'm using is good.
  2. The pump seems to be necessary and irreplaceable. I talked to the mechanical team and replacing the pump with something else will make the project unrealistic.
  3. The 8 hours autonomy is a necessity to the project's success.
  4. The underwater environment.
  5. Dimensions (less than 8 inches long)

QUESTIONS : First, is the 8 hours autonomy doable? What should I do to get that? Change battery? Change pump? Use a totally different approach?


Thank you for your time


3 Answers 3


Lithium Polymer batteries are, in general, the best you're going to get as far as either energy per weight or energy per volume. Going to a different brand may gain your 10% more capacity per volume, but not much more (if you get to that point, and you're in the US, check out ThunderPower batteries -- I fly model airplanes in competition, and the folks I fly against who fly electric either use ThunderPower, or they're not seriously expecting to win).

There's not really a viable alternative -- diesel fuel has about 20 times the energy capacity as LiPo batteries, but you need a diesel engine (which doesn't exist in the form factor you need) or a diesel fuel cell (which even more so doesn't exist in the form factor you need). There's some experimental battery technologies that get a lot of breathless press from University public relations departments, but nothing has been commercialized.

Do all the bits of the fish robot have to work continuously? Can it swim more slowly? Does the pump have to run all the time? If you can conserve power by having it run intermittently, that'll help a lot.

If you can't get a save by operating things intermittently, you need to go back to the mechanical engineers and show them what LiPo cells are capable of, and mention that conservation of energy is a thing. You'll be doing your part by making sure that no power is unnecessarily wasted -- but they have to make sure the goal can be accomplished with the available power.

  • $\begingroup$ I'm in Canada. so i may try ordering from ThunderPower. another challenge i forgot to add in my question is the dimensions (the fish 's dimensions must be less than 9 inches approx 225 mm long). this restrains a bit the choices. $\endgroup$ Commented Dec 30, 2020 at 22:07
  • $\begingroup$ Yes all the bits have to work continuouslyThe robot must check the obstacles and swim in an autonomous way. Yes it must swim slowly so it doesnt affect the fishes in the tank Apparently the use of the pump depends on the density of the prototype if the robot is dense the pump'll work 100% of the time to keep it from going down if not the pump will be only necessary for a vertical movement. Possibility of intermittent use ? actually i'm thinking on adding a power save mode every X minutes or Y hours so that we can double the autonomy $\endgroup$ Commented Dec 30, 2020 at 22:18
  • 3
    $\begingroup$ What is the pump doing? If it's doing something like thrusting upwards for lift, could the fish be designed with something like a swim bladder to maintain neutral buoyancy? Or just designed to have slight positive buoyancy and a place to put ballast? $\endgroup$
    – TimWescott
    Commented Dec 31, 2020 at 17:41

In underwater autonomous vehicles there are basically two solutions to increasing long term autonomy.

The first, better power generation, (which has already been mentioned) won't really work in your case. See here for a discussion page on larger fuel cells

The second, which is more common, is reducing the amount of actuation. One of the most popular underwater vehicle designs is the glider design. The slocum glider which only pumps water in and out and moves a tailfin (IT also has a very low power computer running)

My suggestion: the project in its current state is unfeasible. You need to redesign many components. Just running the motor for 8 hours is infeasible, not to mention the other components you want to run (computers and ultrasounds).


Thank you for your answers and comments. We finally decided to change the pump with a less energy consuming system. We're testing the prototype in the next to weeks we hope we get the estimated autonpmy


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.