We looked at a variety of boards, shields, sensors, and kits of hardware electronic devices. All of them are able to aid and achieve the goal of this project, however, we considered multiple factors to find the best device. The factors are reliability, price, and whether the device is designed specifically for machine learning.
Using the factors, we are able to categorize and prioritize the components from most helpful to least. The sorting assigns importance to each ranking based on the list below:
The list above indicates that “Reliability” is most important, with “For ML?” and “Price” as second and third most important. The table list above is already sorted using Notion’s Table Sort feature.
Based on the sorted list, the most helpful and fit components/devices become apparent to us. The Tiny Machine Learning Kit is the best fit for our situation. It is created by Arduino, which is a reliable company, it is fit for ML and computer vision, since it comes with Arduino Nano 33 BLE Sense and a camera, and it has a low price. Arduino Machine Vision Bundle and Nvidia Jetson Nano Developer Kit are the second and third fit devices for the task.
Comparing the three, we believe that Arduino Machine Vision Bundle will provide the highest outcome for our wildfire detection project. It is specifically created by Arduino for machine learning and computer vision tasks, and it does not have a high cost similar to Nvidia Jetson Nano Developer Kit. With that said, we understand that ~$151 may be high in terms of cost compared to the ~$64 cost of Tiny ML Kit.
What did we learn?
Why is it useful?
How is it useful?
Arduino Tiny ML Kit provides the best package based on performance and cost categories.
With this knowledge, we are now aware of what components we would like to use.
We do not need to spend any more time examining and evaluating other products in the future.