In this month's contribution to the IoT World community site, Ayla Networks' CEO, Dave Friedman, explains what’s behind the smarts in the IoT. The brain in the Internet of Things is more than being tied to the Internet and connecting to an ecosystem of other devices. It's more than a remote monitor for your smartphone or tablet. It's really about data intelligence.

The real value of the Internet has been the capability to provide significant new value in the way we collect, manage and make sense of all the data we create and the Internet of Things is bringing this to our devices.

But to make sense of all this data and truly leverage the potential capabilities, Dave outlines a few keys points to achieve a smarter IoT:

  • Flexible data definition tools: The platform needs to include intuitive tools for the manufacturer to customize the definition for what data to collect for a given device and how it is to be collected (e.g., how often, how much). It also needs to be easy to set up and define new types of devices. There should be no limit on what types of devices can be defined. Manufacturers should have to work with only one platform for all the connected devices they will launch.
  • Data Virtualization and non-SQL database: The system for organizing the device data for retention also has to be very flexible. While accommodating most any data definition of a device, the platform needs to adapt quickly to the fact that device data definitions will change (expand) often, and that each device will have many different data definitions out in the field. As a result, the inherent rigidity of an SQL database -- and the development lag to change SQL database schemas -- is overcome.
  • Data processing and analytics services: An IoT platform really starts to multiply value when it includes services to process the collected data and turn it to usable information. At the foundation are the capabilities to create and control event triggers in a highly customizable fashion. Finally, there are data analytics modules for business intelligence -- e.g., reports and dashboards that the marketing, product strategy, and service support teams each will want to consume.
  • Role-based data access control: Manufacturers need to be able to define various user profiles, each with fully customizable control of what data for which the user has read and/or write access and then to classify each user ID to a defined profile.
  • Data scalability: This may be the least sexy part of the data intelligence potential of the IoT, but it's a critical prerequisite to making it all work. Again, the manufacturer will need good tools to manage the warehouse of data as set grows, e.g., culling and archiving data over the long term. Much of this should be customizable and automated.

You can read Dave's complete blog post here: "Putting the Smarts in the IoT"