Blog entry 6: Project Development - The making of our Chemical device

1. Our team Chemical Device

- In this section, I will briefly describe my team chemical device.

Our chemical device will be an automated CO detection and ventilation system. The system will consists of a CO detector, a fan and vent, which also has a vent door that can open and close according to the CO level to create ventilation, and finally LEDs and a buzzer to warn the user when the CO level in the air is too high.

How the chemical device will work: The CO detector will constantly take readings of CO level in the air. If the CO reading is at safe levels, only one green LED will be turned on, and every other components (fan, vent door, buzzer, red LED) will be off and closed. If the CO reading go above a certain critical value (threshold), this will trigger the Arduino to turn on the red LED, turn on the buzzer for a few seconds to warn anyone in the room, and then turn on the fan, turn on the servo to open the vent door, to create an air flow, to ventilate the room. The vent will remain open and the fan remain running until the CO level is no longer at dangerous levels. When the CO level dropped below the threshold again, the vent door will close and the fan + red LED will be turned off, and then the green LED turned on again, resetting the system to its original state.

- What it is. What problems will the chemical device solve?

In rural areas of Vietnam, especially high-rise mountainous areas, in the winter, many people still utilize charcoal heater to warm their homes.

Each year, there are many cases of people hospitalized or died from CO poisoning, from burning charcoal in closed environments with little to no ventilation. This has been a problem persisting for many years, as people in rural areas have no access to electricity at all, or are unable to pay to have electric cables running to their homes.

Our group aims to develop an affordable, automated CO detection and ventilation system to install in homes. The device will not only warn users of excess CO levels in the air, but also help to ventilate the air when needed, preventing people from suffocating from CO build up in the room from burning charcoal for long periods of time.

- Below is the hand sketch of the chemical device.

First sketch:

After starting our prototyping process, we were informed by Mr chua that we will need some form of mechanism in our prototype, which threw us off our feet, as our original design don't even have any moving parts, excluding the DC motor running the fan. This leads us the adding a vent with a door that can open and close as needed, so we can incorporate the mechanism of gears, more specifically Rack and Pinion, as a mechanism to move the door.

We also received some advices from other people doing their prototypes in the lab on how to improve ours. Their advices turned out to be quite useful: like how to orient the motor to make the fan run smoothly, or how we should use a bigger, thicker shaft to connect the pinion to the servo so it is stable and can withstand more stress.

Due to these new circumstances arising and receiving feedbacks, we redo our sketch to a more realistic, doable prototype that we can execute:

Final sketch:

2. Team Planning, allocation, and execution - In this section, I will list down my team member's name and their respective roles (CEO, CFO, COO, CSO)

• Ngo Van Anh (me) - Chief Executive Officer

• Lim Yan Zhen - Chief Financial Officer

• Jeremy Wong - Chief Operating Officer

• Jeevan - Chief Safety Officer

- I will show the finalized BOM (BILL OF MATERIALS) table.

- I will show the finalized Gantt chart (planned and actual) and the tasks allocation for each team member.

Planned Gantt chart:

Actual Gantt chart:

3. Design and Build Process In this section, I will provide documentation of the design and build process.

Part 1. Design and Build of Rack and Pinion, Door and Stopper, and Cylinder for fan (done by Jeevan). Link to Jeevan's blog: https://axrega.wixsite.com/cp5070-2022-2b02-gro/post/blog-6-project-development

Part 2. Design and Build of Room, stands for DC motor and servo (done by Yan Zhen). Link to Yan Zhen's blog: https://yanzhen21.wixsite.com/cp5070-2022-2b02-gro/post/blog-6-project-development

Part 3. Programming of DC motor, 360 degree servo, LED and Buzzer (done by Jeremy). Link to Jeremy's blog: https://cp5070-2022-2b02-group4-jeremy.blogspot.com/2023/02/project-development-blog-entry.html

Part 4. Integration of all parts and electronics (done by me) - Embed the finalized fusion 360 design files.

- Documentation for integration.

After we finish laser cutting the walls for the room and 3D printed all individual parts.

The 3D printed parts

The laser cut walls for the room housing the prototype

Yan Zhen supervising the laser cutting process

Buzzers that we ordered online

We assemble the system by grouping components into different functions and construct individual functions before installing them:

- Cylinder, motor stand, fan + motor

- Servo, servo stand, rack and pinion

- Door and stopper

- Buzzer does not need to be connected to any other parts beside the Arduino board, hence no assembly needed before installation.

- And lastly, the Breadboard will house the 2 LEDs as well as the CO detector, connected to the Arduino board

The acrylic sheets are glued using acrylic glue to create a room. At first, we do not know that acrylic glue takes a very long time to cure, leading to me and Jeremy holding the acrylic in place for 15 minutes to no results, despite the bottle says it will start to cure after 15 seconds.

We received some advice to use the hot air gun in the soldering station to blast the glued area, helping the glue cure. With the hot air gun, we were able to make the glue set in just 1 minute, and finished constructing the room. The ceiling is left unglued as we still need to install parts as well as make modifications if needed.

For ease of modelling and printing, the bottom part of the motor stand is printed as a square, as we are not sure how to model a rounded bottom to fit the cylinder perfectly. After printing, we sand away a little bit of the bottom and super glue it to the inside of the cylinder.

The cylinder + motor stand is attached to the room using hot glue, the motor and the fan will be mounted onto the stand in the final setup.

The door is glued the the rack in the rack and pinion set up.

Before installing the servo, we wanted to make sure the rack ad pinion system for the door works. So I tried just turning the gear with my hand to see if the door can move smoothly, and to see if there is alot of resistance between the door and the floor.

Fortunately for us, the rack and pinion works quite smoothly, below is a video of me testing them out:

After confirming that our rack and pinion works, I can move onto installing the servo

The servo is placed on top of the stand, secured with hot glue. The pinion is inserted into the servo and fitted into the rack as shown in the picture above. The door lies in between the wall and the stopper, covering the vent hole.

Then, we need to wire the components to the breadboard and the Arduino board. The wires given in the Arduino kit were not long enough, so we needed to solder 2 wires together, as seen in the wires running from the DC motor turning the fan.

Then, we proceed to install the buzzer. This was abit of a last minute decision. Initially, we plan to have the buzzer being just our Arduino board playing a sound. It was not very loud but still sufficient demonstration purposes as an alarm. However, you might be wondering why we have a buzzer in our initial sketch?

This is because we ordered the buzzer from Shopee after we finished the sketch, and the estimated delivery date is 14th Febuary to early March, and that is the fastest delivery we can find. So, we decided we cannot risk waiting for the buzzer and proceed with making the prototype using the Arduino as the "temporary buzzer". Luckily, the Universe and Shopee took pity on us and delivered the buzzer just before we went to the lab to finish up the prototype. So we decided to add the buzzer to stay as close to our sketch as possible.

We forgot to take pictures for the installation of the buzzer, but, it is present in the final hero shot.

In the lab when we finished construction of the prototype, we also forgot to take a picture of us holding the prototype. As we realized this too late, this is the best alternative we can come up with: photoshopping individual pictures onto a picture of the final prototype.

- Hero shot for integration:

Below is our group's hero shot of the CO detection and ventilation system working, with some nice relaxing background music, credited to Aminur from DCHE 2B01.

4. Problems and solutions In this section I will describe the problems encountered in the design and build process and how the team solved them.

Problem 1: Choosing the Mechanism for our prototype

At the start of our designing process, we did not know that we needed a mechanism in our prototype. As it is already too late to order additional parts, we had to make do with what we have. Since we have servos at our disposal, I figured we can have some mechanisms that utilizes a circular movement. A quick brainstorm and we decided we can have a vent with a door that can open and close using a mechanism. We looked into several different mechanisms that can possibly be used to open and close a door: Crank and slider, Cams, Rack and Pinions, Levers.

From our prior experience of doing practical with Gears, we decided Rack and Pinion would be the best and safe option as it is relatively simple and we are already experiences with gears.

Problem 2: Extra 3mm of Acrylic being cut for the room

When we model the walls for the room, we modeled them exact to the 20 x 20 x 30 cm dimensions. We forgot to take into account that the acrylic sheet itself is 3 mm thick, which resulted in a bit of extra width and height when we assemble the room. We tried to counter this as much as possible by placing the walls in a way that minimizes gaps.

However, we do still ended up with some gaps at the top. But since we do not need to actually generate CO in the room the test the prototype, this does not affect the functions of the system much.

Problem 3: Door for the gear powered vent was unstable

Originally when we first thought of adding a vent and door, we did not put any stopper in front of the door. Later we realized this means the door will be very unstable and constantly fall over. We remedy this by 3D printing a stopper, which is just a long rectangular box to put in front of the door, creating a gap between the stopper and the wall where the door can go in between. The stopper held up the door and it stopped toppling over.

Problem 4: The wires were too short

The wires provided in the Arduino kit were too short, so we could not use them to reach some of the components that are higher up on the wall, like the buzzer or the DC motor for the fan.

This is an easy fix as all we need to do is solder 2 wires together to create a wire twice as long, which is more than sufficient to reach all the components. Admittedly, in a bigger prototype, this might prove to be troublesome as we might need to solder more than 2 wires, and the solder connection is not as strong as an unbroken wire.

5. Project Design Files as downloadable files In this section, I will provide all the design files (Fusion360 files, .dxf files, .stl files, arduino programs files) as downloadable files. (upload these files in onedrive or google drive of your personal account. Each person must have these files. Always check that the links to download the files are working.)

1. Rack & Pinion : Rack & Pinion .f3d & .stl

2. Door & Stopper: Door & Stopper .f3d & .stl

3. Cylinder for Fan: Cylinder for Fan .f3d & .stl

4. Servo Stand: Servo Stand .f3d & .stl

5. Motor Stand: Motor Stand .f3d & .stl

6. Wall: Wall .f3d & .dxf

7. Arduino Programming: Arduino Code

And below is the finalized 3D model of the product. As I was unable to find and download the 3D model of the specific CO detector, Buzzer, DC motor, and Servo that we are currently using, I had to substitute them with simple block shapes as modelling them from scratch requires levels of 3D modelling skills far exceed my own, as well as the capability of my laptop which is already freezing every 10 minutes when I try to run Fusion 360.

6. Below is my Learning Reflection on the overall Project Development. This practical was an emotional rollercoaster for me, as we moved through each stage of the design and prototyping process, it seems like we couldn't take 2 steps forward without taking step back. Despite being a very fulfilling experience, if I had known it is this challenging to design and prototype a new product, I would have prepared myself a lot more, in terms of knowledge and skills, as well as in terms of mental resilience.

From the designing process, we already encountered the issue of implementing some form of mechanism into our prototype, which ended up eating up a very large portion of our time and our sanity, because it complicated the code as well as the parts we need to make so much.

I was mostly responsible for coming up with the design of the entire system, and it wasn't easy to figure out a way to incorporate a mechanism into our system without making it too difficult for our team to craft and code. The design I came up with isn't the best, as we constructed it, I realized there are definitely many points which can be improved, like the awkward positioning of the servo, the fact that the door still shifts between the stopper and the wall which can result in jamming a lot of the time, or how the shaft between the servo and the pinion is very flimsy and is literally bending as it pushes on the door. I also had to gave up the idea of making a wall mounted board for the LED as we no longer have as much of a time budget to design and make it now that we have a mechanism to worry about.

I also have to make do with the CO detector attached to the breadboard directly instead of being mounted on the wall, as that would require us to solder wires onto the 3 legs of the CO detector to be able to connect it to the breadboard. And since the gaps between each leg is very small and we only have one single precious CO detector to work with, we couldn't afford to risk ruining it with our soldering.

However, I think the person who suffered the most was our one and only coder, Jeremy. As frustrating as the designing and modelling and 3D printing and assembly is, when I, Jeevan, or Yan Zhen encountered problems, we can seek for advices from other groups doing relatively similar mechanism. Furthermore, Jeevan and Yan Zhen themselves are quite abit more experienced than most of our classmates as they have previously participated in a 3D printing egg drop challenge in NUS. However, our functions are very ambitious compared to other groups, which resulted in the code being much more complicated, so Jeremy couldn't receive as much help from other people in the lab and mostly had to look online.

There are many times when he managed to make one part of the system works, and somehow it resulted in another, previously-functioning part now stopped functioning. Near the end, we were already considering getting rid of some functions, to make it easier for him. But somehow he managed to figure it out on the last day.

Overall, this is a very enriching experience and has taught me a lot about designing and prototyping process.