Fall 2020 Design I Final Competition
Section J: Team Null
Problem Statement: How might we improve adherence to COVID-19 requirements with regards to limited capacity space?
Team Members: Grant Baldwin, Michael Hills, Andrew Smith, Noble Thachiyath, Andy Yu
Instructor: Cara Juergensen
Good work Team Null! Great job getting to the final competition! I have a few questions for you:
1) How do you anticipate getting the information out to individuals? Is this a service for which they sign up? Will this be a free app they download?
2) Do you envision that this device will be positioned at every store? What about multiple entrances and exits?
3) What maintenance do you anticipate with this device? How will it be powered? Can this integrate with existing cameras located at the stores?
4) Who maintains the data and distribution of that data?
Hello, thank you for your questions. Our team worked on these answers together in a group call. Please let us know if you have any additional questions based on our responses.
1) When the device is set up for the first time the user is able to input phone numbers and emails over wifi. This is then used by the program at the appropriate times to send the text messages and emails. Sending emails through the device is free, however depending on the user’s carrier, sending text messages would not be. The exact service used to send text messages costs about 0.02$ per text message.
2) This device would ideally be positioned at smaller stores that currently employ someone to count people entering and exiting the store. Although it sounds niche, there are a surprising number of small stores with single exit/entrances who meet this requirement in Colorado and beyond. Currently, multiple entrances and exits are not supported, as there would have to be a significant amount of setup required to ensure it works properly in each individual location that uses this. The goal of this product is for users to be able to set this up themselves with no professional help required, and having a complicated network of raspberry pi’s interacting with each other to exchange real time accurate data is not something that can be shipped premade.
3) The device does not need any software maintenance but will need to be dusted off every 3- 6 months depending on how dusty the surrounding area where the ACM is installed. The ACM will be powered by a DC converter that is plugged to any 120V wall outlet or could also be powered by a battery, however it is recommended that the user uses a wall outlet. The software that is currently programmed can only be used with the camera that is in the case as it will need a certain angle (pointing down) to work accurately.
4) The user is in full control of all the data. The raspberry pi itself does not store any data, and the software is constantly updating itself and also never stores any data. The only data that is available to be stored are the text messages or the emails that the user gets on the number of people in their store at a specific time, which would be stored on their choice of device. No data is stored, analyzed, or distributed on the raspberry pi or the counting software.
Howdy Team,
A very unique and compelling project idea and great performance results with your execution and real-world validation, effectively conveyed in this concise video.
Questions:
1. Your evidenced real-world performance is impressive, and coverage of a large (wide) single entry/exit point is evident. What would you do for large public facilities with multiple entry/exit points to the building? Is there the possibility, and how would it be executed, of tallying and processing data from multiple units? Could this be used outside or for very large venues? Consider large, urban convention centers, government buildings, amusement parks, etc., for examples. Any modifications or further development required to accommodate these applications involving different environment and multiple units per venue?
2. Is there anything else on the market like this? What are other alternatives other than manual monitoring, counting, and tallying via employees?
3. Though very arguably outside the scope of a first-year course, for cooling purposes, how much heat needs to be dissipated safely from the device, and does any validation (testing and analysis, though you may not have taken thermodynamics yet) buttress the design for this (like your vents, need for active cooling, etc.)?
4. Your works-like prototype is impressive. However, I am not clear as to what the final design will look like and how it will be different from the works-like prototype. Please enlighten us on the differences in features, components, materials, and manufacturing, as you are able, between the device shown here and an envisioned production model.
Thanks so much.
-Prof. Allam
Thank you Dr. Allam for all of your questions. Our team compiled these responses over a conference call. If you have any more questions, please let us know.
1. This concept is viable outside of our design focus however, the Automated Capacity Monitor would require extensive modification. This is mainly due to the fact that the computing system we used, a Raspberry Pi, would not be sufficient to run the software on a large venue. A more powerful computer would be required to monitor the constant flow of people. The device’s cooling would have to be overhauled to an active cooling system, as a passive airflow would not be sufficient to cool the device at this larger scale. If the venue has several entrances and exits all of the upgraded Raspberry Pis would have to connect to a central network which would then provide the user with notifications regarding capacity.
2. Although there are other products on the market that can do what our product does, it is never the focus. Examples are Qudini, V-count, and OpenPath. These other services offer analytics on a subscription-based service, and occupancy tracking is sometimes not offered as visible to the user without paying an additional fee each month (ex. OpenPath has occupancy tracking included only in their premium service). Our system is directly focused on allowing businesses to see how many people are in their store at a given time. The user pays a singular, upfront purchase cost and then has full control of the product in perpetuity. Another major difference our product has is the way that the data is stored. With the other services, you have access to the data but you have no control over what the service provider does with the data. With our product, you have full control over your data, and our system doesn’t collect, send, or analyze a single piece of information that is not critical to the functionality of the product, and the information is directly sent to the user only. No third party, no databases, no information collection system.
3. Simulations that were run were mainly just looking at the airflow temperature and paths, if the surface of the raspberry pi’s processing unit was at 85 degrees Celsius, the maximum temperature it can be before it starts throttling. In the envisioned application of smaller stores, it likely would not run at maximum load for prolonged periods. The chosen vents seemed to have the best ventilation, as opposed to shrouded vents which would have blocked dust from entering the device. A standard Raspberry Pi does not have any devices which provide active cooling, so we did not explore that method for cooling.
4. Our final product would be quite similar in functionality, however, we would change some components to make the design more reliable and easier to mass-produce. For example, something we did not mention due to the video’s time was the mount for the case. The mount was made out of a low-density plastic for our prototype but a final product would be out of a more durable material such as a high-density polycarbonate. We would also like to use plastic molds to produce the case instead of 3D printing, this would be far cheaper and quicker to manufacture. We could use PLA as the prototype used, but it seems plastic for molding (plastic pellets) are much cheaper than 3D printing filament, so other plastics could be explored for manufacturing.
– Any thoughts on a read out or digital display?
– Any design upgrades planned for full implementation?
Thank you for your questions, our team compiled these responses over a conference call. If you have further questions please let us know.
After considering multiple outputs including a digital display, our team decided against using a read-out or digital display. Though some stakeholders prefer a display that shows the number of people in the store, the goal of this product was really to inform the stakeholder when the store was nearing capacity. Implementing a readout or digital display would require a significant amount of work for relatively little gain. If the stakeholder wants, he can receive a text showing the current number of people in the store, but a live-updating digital display is not part of our design.
For full implementation of our design, we would want to change the material for a few key components. Specifically, we would want to change the case and mount to be made out of high-density polycarbonate, instead of low-density plastic. We would also create our parts out of a plastic mold instead of 3d printing, for the final product. By doing so the product will be more sturdy while also being cheaper and easier to mass-produce.