Detailed Design Description: https://coloradoschoolofmines.github.io/wii-track/design
Distributed inventory tracking using WiiFit, Rasberry Pi, and Arduino. Backed by the computing power of AWS Lambda, and the flexibility of the DynamoDB NoSQL database.
This project won the following prizes at the HackCU hackathon.
- Judges Favorite
- Dish Network - Asset Tracking Award
- Amazon Web Services - Best Use of AWS
Currently, inventory tracking is only done at centralized locations. This leaves many remote areas with limited inventory tracking capabilities. Our project is designed to solve this problem by enabling distributed inventory tracking.
Wii-Track uses weight and color sensors to identify objects.
The status of the entire system can be monitored from a simple user interface. Supervisors can view pictures of packages and remotely resolve any issues that may arise.
The power of our design is that all of the computationally intense data analytics and image processing take place on AWS Lambda. This allows the remote inventory tracking nodes to be very lightweight. We used a Raspberry Pi, Arduino, and WiiFit for this implementation, but we could also use a low cost embedded device and produce the same effect.
We use AWS DynamoDB to track the inventory. The desktop user interface allows users to query this database.
We are extremely inexperienced with AWS. As such, it proved difficult to navigate the AWS portal.
- We were able to successfully utilize AWS Lambda and AWS DynamoDB. We had never used these technologies before, so it was a huge learning experience for us.
- We were able to successfully connect a Raspberry Pi, Raspberry Pi camera and Arduino to create remote image recognition station.
- We were able to utilize the WiiFit board to collect weight data on inventory items.
We learned how to use many exciting technologies including AWS Lambda, AWS DynamoDB, and the WiiFit API.
Wii-Track is designed to be highly extensible. Additional edge nodes can be added easily, and more complex analytics can be integrated into the AWS Lambda function. Additionally, Wii-Track can utilize different hardware for measuring characteristics of the inventory. For example, we could use infrared sensors in addition to the camera and scale that we utilize today. We can also enhance the supervisor portal for viewing packages remotely.