Analytics of Things (AoT) – Considerations

Analytics of Things (AoT) refers to real time analytics and batch processing of sensor data generated by Internet of Things (IoT) devices. If you are considering an AoT platform, here are the things you should consider.

  1. Massive Scalability: Sensors produce massive amounts of data. Hence, your AoT platform or application must be massively scalable. Please refer to Massively Scalable Applications presentation as scalability is of utmost priority for AoT applications.
  2. Machine Learning/AI: Business Insights are useful. They help businesses hack growth. Algorithms today can crunch numbers and big data, they can self learn, interpret patterns and come up with business insights. Consider how much Machine learning or AI capabilities do you need in your solution?

In addition, you should consider the following questions:

  1. For what duration do we need to store the live stream data? Days, months or years? Sensors produce massive amounts of data. 1 million messages per second fill 1 GB of your storage in a second (assuming 1 message to be of 1KB for simplicity). That fills your 1 TBĀ PCIe flash card (or other hard drive) in just 17 minutes. Though it will be nice to be able to show sensor data in live graphs but limiting data stored can save a ton on hosting fees.
  2. Can sensors send data over plain TCP protocol? Can we avoid the unnecessary overhead of HTTP or REST? How important it is have a web service (HTTP/REST) receive the data?

TechFerry has been involved in the process of helping many clients with AoT, Big data analytics, machine learning, IoT and massively scalable applications in general. If you have questions, please feel free to Talk to TechFerry.

This entry was posted in AoT, IoT and tagged , . Bookmark the permalink.

Leave a Reply

Your email address will not be published. Required fields are marked *