Our hardware team performed multiple design cycles on the device prototype. The general design cycle for all the parts usually consists of collecting data (for example, taking measurements and researching the material to use), making the initial prototype, testing the prototype with multiple users, assessing the prototype’s usability, and modifying the prototype based on the testing and user feedback. The threads fitting and mechanism designed for the DNA extraction tool and liquid handling process were our main contributions to achieving safe, functional, intuitive products for onsite PCR experiments.
We also demonstrated engineering success in the wetlab and dry lab team in primer design. Our prediction model is able to rank primers according to the criteria. Its accuracy validated by the wet lab as our primer sets with low confidence ranking have a failure in amplification. To build upon our engineering success, we can apply machine learning and build our model with a larger dataset.
Extraction Tool
Iteration | Design and Build | Learnings from testing the Design |
---|---|---|
1st Generation (Fig. 1) | 3D-printed drill design has the conventional screwdriver mechanism. | - Able to drill out wood dust (granular) from the sapwood layer. - Hard to apply hand force and the drill moved around when forces was applied. |
2nd Generation (Fig. 1) | A drill handle (similar to the corkscrew) was added to the 1st Generation design which allowed users to apply more power when drilling. | - Able to apply force better and without moving the drill during the extraction process - Able to obtain wood dust form the sapwood layer, however the handle is too short. |
3rd Generation (Figure 3) | Same concept from 2nd Generation, however the drill handle size increases and finger slots were added | - Successfully obtain wood dust from the sapwood layer - The handle size and finger slots fit perfectly to grasp the drill - Able to apply force without moving the drill during the extraction process. |
Figure 1: 1st Generation Extraction Tool |
Figure 2: 2nd Generation Extraction Tool |
Figure 3: 3rd Generation Extraction Tool |
---|
Liquid Handling and Mixing:
To validate the liquid mixing mechanism we used a pipette to control the input and output amounts of water liquid. After collecting 6 sample points, we observed 14.3% error on average with standard deviation of 1.2%.
Iteration | Design and Build | Learnings from testing the design |
---|---|---|
1st Generation (Figure 4) | Using Scotch Yoke Mechanism, rotational motion translated to vertical movement. The user rotate the handle and move the cylinder up and down to crush and mix the liquid in the container. | The design could mix the liquid however it cannot perform grinding the sapwood to small pieces to extract the DNA. |
2nd Generation (Figure 5) | Similar to garden hose valve with an opening valve inside. | 3D printer could not print the design and it has no foil cutter. |
3rd Generation (Figure 6) | Foil cutter was added as a needle. Rubber bands and rubber band slots hold together the three-piece design. | Needle was not easily printable from 3D printer. |
5th Generation (Figure 8) | The connector tubes were redesigned for a smaller liquid container. The foil cutter’s size was redesigned as well. The rubber band design was removed | Dimensions were off. The rubber band design was removed because the tube held together well enough and the rubber band complicated the design. |
6th Generation (Figure 9) | The design (including the threads) now fits the liquid container. | Dimensions were off as shown in the figure 9. |
7th Generation (Figure 10) | A new cutter design was used because of the smaller liquid containers. | Too small to print. |
8th Generation - Current Design (Figure 11) | The cutter was enlarged. The aligning mechanism was also redesigned to make sure it was printed properly. | Worked successfully and printed properly |
Test results with current design - 8th Generation To validate the liquid mixing mechanism we used a pipette to control the input and output amounts of water liquid. After collecting 6 sample points, we observed 14.3% error on average with standard deviation of 1.2%.
Initial | Amount Added | Amount Final | Amount Error % |
---|---|---|---|
100µL | 400µL | 440µL | 12% |
100µL | 400µL | 420µL | 16% |
100µL | 400µL | 430µL | 14% |
200µL | 300µL | 433µL | 14.4% |
200µL | 300µL | 427µL | 14.6% |
200µL | 300µL | 425µL | 15% |
Figure 4: First generation of the liquid mixing and handling design
Figure 5: Second Generation of the liquid mixing and handling design
Figure 6: Third Generation of the liquid mixing and handling design
Figure 7: Fourth Generation of the liquid mixing and handling design
Figure 8: Fifth Generation of the liquid mixing and handling design
Figure 9: Sixth Generation of the liquid mixing and handling design
Figure 10: Seventh Generation of the liquid mixing and handling design
Overall goal, Current Step, and Next Step:
Based on the requirement to perform the LAMP (PCR) reaction on site, we attempted to design a portable heating unit that can heat up to 71C (+ or -1C) consistently for 30 min. We currently have a prototype that is portable and can sustain heat up to 100C. We also found a suitable thermistor that measures the temperature with a one-degree fluctuation. The main components we used included a PTC heating module, a 10k thermistor, an Arduino nano, a rotary encoder, and an OLED screen for readings. For the next step, we plan to work on a consistent heating function for a low-cost and portable heating unit. We also aim to design a PID control algorithm for precision, a portable battery or alternative battery options for forestry practitioners and researchers to work on site.
Based on the requirement to perform the LAMP (PCR) reaction on site, we attempted to design a portable heating unit that can heat up to 71C (+ or -1C) consistently for 30 min. We currently have a prototype that is portable and can sustain heat up to 100C. We also found a suitable thermistor that measures the temperature with a one-degree fluctuation. The main components we used included a PTC heating module, a 10k thermistor, an Arduino nano, a rotary encoder, and an OLED screen for readings. For the next step, we plan to work on a consistent heating function for a low-cost and portable heating unit. We also aim to design a PID control algorithm for precision, a portable battery or alternative battery options for forestry practitioners and researchers to work on site.
We tested Peltier, PTC, and coiled heaters and selected the PTC heater as it consumes the least amount of energy with room-efficient build for portable on-site heating purposes.
We tested a 1K NTC thermistor, a 10K NTC thermistor, an IR temperature sensor, and a 5V 10A temperature control relay. The 10K NTC thermistor has the most accurate temperature reading since it employs contact sensing.
Then, we tested with the temperature module. We needed more control of the interface and indicators for practitioners to understand.
To address this issue, we designed a LCD screen that can display input reading.
We also have designed the heating plate made of aluminum for the LAMP reaction
(Version 1.0 on the left and version 2.0 on the right)