This is our design process for the hardware. Our overall aim is to design a device that places engineering yeasts and achieve the following objectives:
· Receiving signal molecules that are distributed from tea plants attacked by aphids, and passing them to the specific olfactory receptors on the yeasts
· Providing adequate nutrition and appropriate condition for engineering yeasts so that they work properly and stably
· Making the sex pheromone molecules produced by engineering yeasts quickly spread into the air
· Convenient for manual adjustments when necessary

Our initial design was a portable box-shaped device that could hang on the branches of the tea plants. Filter membranes that can prevent outer bacteria from invading are installed on the "windows" of the box connecting inside and out so that air circulation could be guaranteed. The fluid medium is placed in the bottom of the box with engineering yeasts in it. Random shaking culture is possible depending on the environment (wind, shaking of the branches, etc.). If tea is planted open-air, a rain shield is added above the device to avoid special weather conditions. Small lids are installed inside the box around the side wall to prevent the fluid medium from spilling out in extreme conditions.

When it comes to materials, MCE, PP, PVDF, and PTFE are the ones we found suitable for our anti-bacterial air filter membrane. After comparing their prices as well as effects (air permeability, bacteria prevention, etc.), we selected MCE, which is commonly used in filter columns in laboratories and has promising effects, as the material for filter membranes.

Apparent drawbacks can be seen in the first model as listed below:
1. The signal molecules released by tea plants are mainly hydrophobic. In this model, volatile molecules should be dissolved in the hydrophilic fluid medium before it is detected by engineering yeasts. However, since such signal molecules from plants may only exist with the concentration within ppm level in the air, detecting them in this way by yeasts is nearly impossible. Moreover, if engineering yeasts could detect these molecules, the hydrophobic sex pheromone molecules could be blocked by the fluid medium and thus have smaller possibilities to be distributed out of water. 2. By placing the box with fluid medium the following risks could be brought:
①Fluid medium spilling out with engineering yeasts in extreme conditions, causing pollution in the tea farm which is bad for biosecurity;
②Probabilities for tiny raindrops to enter the device thus cause pollutions for growth environment for yeasts (with membranes on, the possibility could be reduced).
③Expecting random shaking is not a realistic approach. In most cases, the yeast would just be at the bottom of the fluid medium without adequate shaking. So coping with the problem the liquid medium brings becomes one of our major issues.

Concerned about the drawbacks mentioned above, we had the 2nd plan:
The 2nd plan aims to "improve the chances of contacting with air for the engineering yeasts". We decided to reduce the depth of the fluid medium and expand the contact area between the fluid medium and air. 1. Based on the 1st plan, the bottom of the box is changed into uneven surfaces with tiny bumps, and the volume of the fluid medium is controlled to a smaller extent, making a thin layer of liquid "coating" on the surface. Thus, the possibility of yeasts on the "peaks" having direct contact with air is improved. And with random shaking from the environment or manual methods, the yeasts can also be "washed" down from "peak" to "valley" to ensure the normal nutrition supply. 2. Based on the plan above, we had another tentative 2.5th plan with the same bottom surface change, but the material of the bottom surface is adjusted so that engineering yeasts could be immobilized on. A similar volume of the fluid medium is used to provide immobilized yeasts with conditions for basic growth.

Though improvements have been made to our 2nd plan on the chances for yeasts to contact with air, the fluid medium is still a barrier for yeasts to receive signal molecules from the air and brings plenty of uncertainty like vaporizing, etc. Hence, we come up with the 3rd plan.

The 3rd plan is our attempt to get rid of the "liquid" structure of the culture medium. Here is the rough model we come up with:
Considering immobilizing yeasts on the material that maintains yeasts' growth and allows air to pass at the same time. We plan to make a solid medium panel that has a loose structure on which engineering yeasts are immobilized. The panel itself contains nutrient substances for basic growth and visible small holes for air circulation. When needed, only replacing the panels is necessary. The device is a box resembling the initial design and containing the medium panels, which will be standing among the tea plants, like what the slime boards do. Then it would be more possible for yeasts to capture the signal molecules from the air and release the sex pheromones without barriers. Another plan is to put alginate beads (on which the engineering yeasts are immobilized) in the narrow string bags based on the panels in the idea above, with the inspiration from the yeast biosensors underwater. However, it may not support long-term detection and thus needs further study, since the time of pest attacks could be uncertain.

[1] Wikol M, Hartmann B, Brendle J, et al. Expanded polytetrafluoroethylene membranes and their applications[M]
//Filtration and purification in the biopharmaceutical industry. CRC Press, 2007: 643-664.
[2] Sweetman L J, Alcock L J, McArthur J D, et al. Bacterial filtration using carbon nanotube/antibiotic buckypaper
membranes[J]. Journal of Nanomaterials, 2013, 2013.
[3] Vopálenská I, Váchová L, Palková Z. New biosensor for detection of copper ions in water based on immobilized
genetically modified yeast cells[J]. Biosensors and Bioelectronics, 2015, 72: 160-167.
[4] Miller R A, Lee S, Fridmanski E J, et al. “Scentsor”: A Whole-Cell Yeast Biosensor with an Olfactory Reporter for
Low-Cost and Equipment-Free Detection of Pharmaceuticals[J]. ACS sensors, 2020, 5(10): 3025-3030.