IOT Plant Floor Sample Architecture
Architecture overview IOT Device – Analytics: Gus Segura
There is more than on way to implement IOT analytics for the Plant Floor. For me, its all about guaranteed message delivery (from the IOT Device), micro-services vs monolithic architecture and edge analytics (de-coupled from the cloud). You have a choice on how you implement an IOT Platform. The diagram and quick article is a brief overview of what to expect to find in any IOT Platform Architecture. This is typically a production floor with 1000’s of sensors and millions of messages flowing through the system at any time.
Plant Floor Production, Re-Work, Quality, Delays cost Money!
- Plant Floor should be sending messages to the edge store and the Cloud.
- Monolithic Systems are giving way to Micro-Services – It Just makes sense.
- Real-time data pipelines enable machine learning, visualization and process – now!
- There is plenty of room for Big Data (Hadoop, MongoDB, GraphDB) and Relational.
- Docker-ize everything that makes sense – including the analytics, deep learning, visuals.
- There are many ways to move data around after its in a data lake.
- Consider having a strategy that includes more than one path to your action window.
- Data Discovery – enabling the data sets for analysis, machine learning, visualization
- Make sure you have a tool and/or process so that stake-holders can find the data that is actually being generated.
- Data Quality will keep you up a night
- Unless you implement a strategy for sample, correction and re-work.
- A Data Engineer is not the same as SDE
- However- DevOps can and does work very well for Data Engineering and Data Pipe-line.
If you’re having problems moving to this paradigm, Please give me a call or email me Today.
Do you need help with your IOT Data Project- Please Contact US.
Email: support@blueskymetrics.com – Phone: 765.325.8373 ( call / text ).
Stay Informed: Please Contact US | Subscribe to our Blog Updates or If you would like more information.