Nii-Okaitey Tetteh, Chi Nguyen, Likhitha Chintareddy | Carnegie Mellon ECE Capstone, Spring 2023
This week I received the Jetson Nano 2GB and the Sainsmart IMX129 camera module that I requested from the course list. However, upon receiving the box, I realized the box did not have a power cord or SD card to boot up the OS, so …
Half of the week was spent on the Design Review. There were issues related to my part specifically that were addressed in our feedbacks. First of all, there was a concern related to how we would customize the Yolov7 model to account for human objects …
This week I researched on what is the optimal database for us to use. We considered DocumentDB by Amazon AWS as it is a popular choice as a cloud database, however, MongoDB Atlas emerged as the better choice. As discussed in our presentation last week, we intend to use MongoDB as it allows for the dynamism of our data. We needed to find a way to store the data processed by the software backend and accessible by the frontend. MongoDB Atlas emerged as the better choice based on our criteria that prioritized pricing, portability and features. MongoDB Atlas is free, for starters, while DocumentDB has a limited free tier. Atlas is also Mongo’s native product with a lot more compatible features that seem to make it a better deployment option, as explained here: https://www.mongodb.com/compare/documentdb-vs-mongodb. I have been studying on how to use MongoDB in conjunction with Django as my experience so far has been with mySQL which is a much simpler database. Orthogonal to this, I have begun work on setting up the local Django app environment.
Some changes have been made to the schedule to reflect the work that has had to be put in to find a good database. This week was spent on setting up work for the Figma UI prototype. The upcoming week will be spent on finalizing this prototype and setting up the app using Django framework. The classes that gave/will give me the requisite knowledge are an amalgam of HCI/Design classes and Web Application Development.
Referring back to our team status report from last week, we mentioned that not having our design flow fleshed out before the design review was a significant risk that could jeopardize the success of the project. This issue has been resolved as we spent a …
As mentioned in my previous status report and in our schedule, I have been working on implementing YOLO algorithm for object detection process. I have created a github which includes all of my code and notes on the design and implementation process (you can find …
This week I worked on revising our hardware plan for our project. After speaking with course staff, it seems given our application that an FPGA may not be the best solution for us. Previous projects that performed similar functions to ours utilized an actual embedded computer focused on AI acceleration. Specifically, they used the Jetson Nano. I have never used this board before, so before we get too far into implementation I wanted to familiarize myself with some of the board set up. I watched some YouTube videos about the setup of the Jetson Nano, as well as some example projects to get a feel for how programs are ran on this board. Here are the videos I watched:
Additionally, because we are using a Jetson Nano instead of an FPGA, the way we had hardware sending its information to the web application also had to change. Thus, after researching past CV applications, it looked like a common way of sending information from the board to the web application is through JSON over wifi. I have never done any JSON applications either so I watched an introductory video to get some basics:
Currently I am still on schedule. I requested both the board and the cameras we are planning to use from the course list, so hopefully I can receive them next week and begin setting things up.
That being said, as the design review is this week, I hope to receive some final feedback on our design approach, and start implementing it this week. Specifically, I would like to have the board initialized by the end of this week.
ECE courses that cover the engineering science and mathematics principles that we used to develop our design: 18-349 Introduction to Embedded Systems
The second principle would be writing JSON scripts, and I began learning this principle through the video listed above.
For this week, I’ve been spending time on the design aspect of my algorithm before actually implementing them next week. The goal for this following week will be to implement the YOLO algorithm for object detection and have it work on images with 1 to …
This week, I started doing research on the different UIs used by apps and also published papers and other work in the UI space on how to design for users who are looking at wait times. Some of the work I have found useful and …
The most significant risks that could jeopardize the success of the project are us not having our design flow fleshed out before the design review. This is why this week we focused on sub portions of each of our tasks. By doing this, when we integrate all our design elements together the system should be more fleshed out than before. This also helps us with contingency plans. Now that each of us are getting a better grasp of the inner implementations of our parts, we can all start thinking of contingency plans specific to our implementations.
No changes were made to the existing design of the system.
Our project includes considerations for health and welfare. The issue goes beyond academia – in a capitalistic system where productivity is often maximised at all costs, including compromising on workers’ health and wellbeing – giving people the ability to make informed choices about eating is an important undertaking. Studies have indicated that for a lot of young people, college years are when they begin struggling with eating disorders. Busy schedules and disordered eating can be a contributing factor. And by being able to see what options they have in the time they have to eat could potentially help mitigate that.