IoT is not a solved problem.
What it does
This project provides as much data as a Data Scientist/Analyst could ever want.
We put together a software and hardware framework to enable AI/ML practitioners with the data they need to provide actionable events and notifications.
The Common Sense approach is to display competency in harvesting data from transducers connected on (theoretically) any bus that is available from a synchronous or asynchronous basis.
How we built it
We somewhat recycled a previous project with some refactoring, added features and a more generalized approach to moving signals from transducers and buses.
We created libraries and daemons for consuming data from I2S (audio), CAN and I2C buses and then used MQTT clients and brokers to move the data to downstream platforms. The platforms we used are an on-premise Jupyter notebook and the Exosite ExoSense platform.
We wanted to show how data can be dealt with at the edge to enable actions and insights from various analytics and insights platforms.
Challenges we ran into
Accomplishments that we're proud of
What we learned
What's next for Common Sense
- More buses
- Mainflux integration
- AWS/GCP integrations
- A board that provides a smattering of sensors that utilize our supported buses
- Find partners for agriculture and industrial applications.