The hardware component of this interdisciplinary project uses various engineering principles. We started by consulting researchers and research papers to find issues that we can solve in relation to harmful algae blooms. We found many issues, but after weeks of feedback from people around the community, we locked into something that can be applied to not only our project, but also to other researchers. Our team started with a rigorous prototyping process and moved to an iterative design process until we got to the complete product. All the while, we had one major goal in mind—to create an integrated system that can be inexpensive and usable by anyone and everyone.
Cornell iGEM’s reHAB simplifies data collection and provides a way to use data insightfully. Our system is composed of hardware components and biological components. Our integrated system begins with HabCab. It is an automated inexpensive sampler which will collect samples in water along with storing the GPS location of the sample collected. Next, once we find out that the water contains microcystin toxins, we can utilize the HabLab, a bioreactor that utilizes a flow dispersion nozzle to optimize the flow of water through a microcystin-removing array of alginate beads. Our wet lab team designed the bacteria responsible for the breaking down of microcystins which are locked safely inside each alginate bead.
Our sampling system consists of one vacuum pump connected to a solenoid valve array which regulate water inflow of each sample through plastic tubing. This apparatus can be placed on any aquatic vehicle, and in our case we built a small boat. The gps module on top of the boat coordinates the movement of the boat, the vacuum pump, and solenoid. Once we arrive at each location, an unused tube’s solenoid valve opens. An arduino controls the entire process.The vacuum pump transfers the water through the open valve into the test tube to complete a sample collection. After 8 seconds, the pump turns off and the valve closes. This specific valve remains closed for the duration of the trip. The number of seconds that we keep the pump on for and valve open for was determined through trial and error to determine how long it would take the test tube to fill up without overflowing.
By using one pump for all of the samples, we sped up the design process and made it much cheaper to build. However, contamination became a much bigger problem. Because of this, in order to use different tubes all connected to the same pump and only have one fill with water at a time, we incorporated the solenoid valves into our design.
We modeled a prototype of the sampler vessel in Fusion 360 before fabrication. We then constructed the vessel by laser-cutting acrylic sheets and connecting the pieces with a solvent weld. We chose the method of laser-cutting because it is precise and can cut the acrylic sheet material with ease. The solvent welding method was extremely effective in creating strong, waterproof joints in the boat, which are necessary to prevent leaking. We also 3D modeled and printed a holder for the sample collection device using ABS plastic. We chose this additive manufacturing technique over a subtractive technique due to its accessibility and low cost of materials. Since the holder is not under considerable stress, the ABS plastic material is sufficiently strong.
The sampler vessel is a 3 foot long radio-controlled electric boat. We constructed the hull using 0.08” thick acrylic sheet plexiglass, joined together with a solvent weld in order to create a watertight joint. We laser-cut the pieces from large acrylic sheets to allow for a precise fit. We coated the hull in rubber sealant in order to further protect against leaks. An 11.1V lithium-polymer battery powers the boat’s electronics, including the sampler, microcontroller, and two electric motors, and allows the boat to run continuously for approximately 15 minutes before recharging. We chose a lithium-polymer battery for its large power-to-weight ratio and its ability to sustain the high level of current needed to drive the vessel’s many electronic components. The vessel can be controlled via a Turnigy radio frequency transmitter/receiver system. We started with a more complicated boat design but decided to move towards a more basic design so we can focus on building the other components.
In order to propel the B.O.A.T., we used two NQD757 boat motors wired in parallel to the 11.1V 3S LiPo battery. To generate thrust, these twin motors rely on a propeller to take in water from under the boat and expel it out the back. Because the motors are not effective at steering the boat on their own, we attached a simple servo that controls the direction of propulsion. In order to accurately record the location of the samples, we also incorporated a NEO-6M GPS module into our design. This device also allows the potential to implement a waypoint tracking algorithm that enables the B.O.A.T. to collect samples at user-defined GPS coordinates.
Our current sampler mechanism is capable of collecting up to four water samples from various locations around a lake. It consists of a vacuum pump, pneumatic solenoids, tubing, and four sample collection vessels. Four rubber tubes, which are connected to the sample collection vessels, extend over the sides of the boat and into the water. Between the sample collection vessels and the vacuum pump, we installed cellulose filters that permit airflow into the vacuum but block waterflow, in order to protect the sensitive electronics in the vacuum and solenoids. When a signal is sent to collect the first sample, the corresponding solenoid, which acts as a gate that can open and close a tubing channel, opens. The vacuum pump is then activated, and water from the lake flows up one tube into one of the collection vessels. When the vessel is full, the vacuum pump shuts off, the solenoid closes; the boat is now free to move to the next sampling location. This process can be repeated for up to a total of four samples. The GPS location, time, and date of each sample is recorded, and can be accessed later on dry land so that patterns in algal blooms can be analyzed.
In order to integrate the various electronic components of our BOAT into a cohesive module, we used an Arduino Uno board. To prevent voltage and current overload, the Arduino was powered from the LiPo battery through a 5V UBEC (Universal Battery Eliminator Circuit). Furthermore, because the Arduino can only output up to 5V at 40 mA of current, we needed to incorporate a MOSFET array that enables the microcontroller to regulate the voltage between the battery and each of the 12V solenoids and vacuum. Transistors were chosen instead of relays due to their smaller size, cost-effectiveness, and potential to be scaled down in future iterations. To activate the solenoids, the Arduino interprets the signal from the RC receiver, which arrives as a series of digital pulses. Each time the switch is flipped, it opens a different solenoid and turns on the vacuum for a predetermined amount of time; this process is repeated until the sampling capacity of the B.O.A.T. has been met. In order to power the two driving motors, we wired them to the battery through a Spark motor controller connected to the RC receiver via a PWM cable. Because the spark controllers and RC receiver operate using the same signal protocol, we were able to circumvent the Arduino and directly connect the two components. To operate the motors, the Spark interprets the RC signal corresponding to a selected joystick axis and maps it to a corresponding motor power. Steering works in a similar way; it can also directly translate the joystick reading into a precise angle of rotation. The GPS communicates via a software serial interface and its coordinates can be used to control the steering if autonomous mode is activated.
The bioreactor consists of two sizes of inert tubing ¼ inch radius and ½ inch radius, flow distribution nozzles, 2 filters and a syringe. The water sample to be cleaned starts in the syringe and is pressed through the ¼ inch into the main compartment of the bioreactor. The main compartment consists of the ½ inch tubing filled with our alginate bead encapsulated bacteria. The water continues moving through, interacting with the bacteria so that the toxin breaks down, and then exits the system fully cleaned on the other side. Inert tubing was chosen in order to prevent chemical reactions with the microcystins or any other molecules in the water. Filters were included on both ends of the bioreactor in order to ensure that bacteria did not travel outside of the reaction vessel, even if they somehow escaped the alginate beads. Nozzles were included in order to evenly spread out the flow from the small tube coming from the syringe to the large reaction vessel tube. We designed the nozzles using Autodesk Inventor and then 3D printed them in our lab.
We used a MakerBot Pro to 3D print the bioreactor nozzles out of ABS plastic. This method was the simplest way to produce the small and intricate parts. A subtractive manufacturing method wouldn’t have been able to accurately produce the interior chambers necessary in the design of the nozzle. The tight 3D printer tolerances resulted in fully watertight pieces.
In order to maximize the efficiency of the bioreactor, we had to engineer a nozzle to ensure that water flow is distributed evenly throughout the bioreactor chamber. Because the reaction yield is contingent upon how many beads are in contact with the contaminated water, it is critical that the flow distribution generated by the nozzle minimizes the occurrence of fluidic dead zones where there is no flow between the beads. To achieve this, we used a high-resolution 3D printer to create a cone-shaped nozzle with many small openings at numerous angles. When the water is pushed through the nozzle, it is forced out these openings and distributes itself evenly over the alginate beads inside.
