Open Source Underwater Distributed Sensor Network

There are many concerns regarding what materials we choose to use for the clams. We have to assume that they will be lost and that animals will try to eat them. Additionally, an eventual goal is to develop clams that are benthic and can perform long-term missions. Instead of staying planktonic, the clams would settle out of the water column onto the seafloor, and either take measurements where they are situated, or periodically ascend to take measurements. Therefore, even in normal operation there is the potential for plastics to leach into the environment.

For the initial prints, I believe that we should focus on using polylactic acid (PLA). It is a common filament for 3D printers and it is naturally derived and biodegradable. As of yet, I do not know it’s seaworthiness, but that should be a focus of testing. If anyone has a fish tank, saltwater and a PLA print that they are willing to sacrifice, I would be grateful for their observations about how well the PLA stands up over time. Future testing could use the sensors we incorporate into the clam to measure leaching, but for right now, sight and touch would be good enough to measure seaworthiness.

In addition to knowing how well it stands up to saltwater, we need to know how well PLA performs with biofouling. Biofouling is a serious concern for water quality instruments. It is especially concerning for 3D prints, which tend to have significant gaps and crevices on their surfaces. This is another area where we will have to perform extensive testing.

A common way for mitigating biofouling is to clean water quality instruments after use. It would be too laborious to require researchers to clean each clam individually, so we should try to design them to be cleaned in the dishwasher. As far as I know, PLA softens in dishwashers to a degree that might not be acceptable. We could try to design the shells, so that they can soften without losing their shape or impairing their function. This would also require testing.

If we tried to use a different plastic that is higher temperature, we might lose the advantages of having a naturally derived and biodegradable plastic. A thin conformal coating of silicone or another high temperature plastic might be a good compromise that could fill the gaps to prevent biofouling, and protect the shells from melting and deforming.

Alternatively, we could borrow a solution from natural clams and use calcium carbonate. I have done some experiments with mixing plastics (PLA and epoxy) with calcium carbonate, and it is a wonderful material. It feels like bone and shell, and it is very strong (here are links to vegan taxidermy I made with the material: antlers, scrimshaw).

PLA and calcium carbonate filament is available, so using that material might be a good area to explore. Using calcium carbonate would not add any unnatural chemicals, and I believe that it might help with the dishwasher issue. Also, it might allow the clams to degrade somewhat gracefully, especially if we could figure out a way to build up the shells primarily in calcium carbonate with just a little PLA binder.

This project is inspired by GNU, RepRap, and the free software/open source movement. I am using FreeCAD, KiCAD, and the GNU toolchain for my work. I will be setting up a public git repository, in order to share and collaborate. I am grateful for any criticism or help. In particular, right now, if there are any software developers who are skilled in DevOps, I would appreciate their insights.

One of the big challenges for this project will be testing the clams and incorporating the feedback into the design. We need to lower the bar for participation, so that an individual without expert knowledge can print out and assemble a clam, test it, and then submit the results in a way that can be easily used by developers.

Plastics

For the initial experiments, I used blown acrylic sheet and tubing to make the clam body. The shells were blown using a plastics blow oven, and the internal structures were blown using a heat gun. The seals were made of silicone, the bladder was made of latex, and the different structures were assembled with epoxy.

I used acrylic because it allows for a more direct process than prototyping with 3D printing. Ideas can be quickly tried out and modified with a bandsaw, sander, heat gun and hot knife, instead of working in a CAD file and printing new prototypes.

Electronics

The clam's circuit is composed of a bluetooth module, atmega, push-pull circuits, dc-dc converters, batteries and a wireless charger. A thermistor was used as a stand-in for other sensors, and data-logging was not implemented. This circuit was enough for the clam to control the solenoids and communicate data between the clam and a bluetooth device.

After breadboarding the circuit, I free-form soldered the electronics and gave it a conformal coating of epoxy. Like free-blowing plastic, free-form soldering is an intuitive and direct way to prototype. Different circuits can be easily explored, instead of going through the process of designing and fabricating new PCBs.

Bivalve

In my experimentation, I was able to work out the design of the shell, hinge, and adductor to form the bivalve. The hinge is formed by teeth, like an exaggerated version of the hinge teeth of natural clams. It has two fixed states (opened and closed) that are held in position by permanent magnets. The solenoid for the adductor is only used to change state.

In the closed state, the clam is sealed. In the open state, the clam has a gap to allow gas to release and water to flow in. Right now, that water inflow is what is used for sampling. It may be necessary to add a siphon to pump water into the clam for the sensors.

Bladder

I experimented with chemical reactions to fill the bladder. I did not want to use a pressure chamber to store the gas, because I did not want to build something to maintain those forces, or have a consumable component like a CO2 cartridge. I experimented with acetic acid and sodium bicarbonate (vinegar and baking soda), as well as hydrogen peroxide and catalase (from yeast), and both worked well. My only concerns are that the chemicals are nontoxic, do not affect the sensors, and are easy to clean and replenish.

Battery

The initial idea was for the clams to have copper on one shell and zinc on another to make a salt water battery. I scrapped that idea because the batteries I built lacked power, and an open battery could affect sensor measurements. I decided to use lithium batteries for now because they are powerful and can be recharged.

If a clam is lost and not recovered, then the lithium battery would be a pollutant, so a different battery would need to be used in the future if recoverability becomes a problem. A better salt water battery could be made that was more powerful and closed, so that it would not affect the sensors. In operation, when the clams are first seeded into the water, an opening in the battery would draw in salt water for use as an electrolyte. The opening would close and the clam would use that battery for it's data-logging mission. The batteries would not be able to be recharged electrically, but the clams could retain their wireless coils, and be powered wirelessly during data recovery. The salt water battery could be recharged for later missions by adding more elemental copper and zinc.

Now that the initial ideas have been explored, the design process will shift to working in CAD files for 3D printing and PCB fabrication.