The IoT Ecosystem

K2’s Software Defined IoT platform allows Equipment Manufacturers to build IoT applications that can not only Monitor, Analyze, Visualize data on a real time basis but also provide Controllable Actions which make use of that data


Why Data Analytics

Data Analytics the traditional way

  • Volume: IoT data is enormous in quantity. A lot of these data are redundant, but at the same time mission-critical information is embedded into vast data. Thus, the platform needs to be smart enough to know what to look for.
  • Cost: The cost of sending IoT data to the cloud could be very expensive. This includes cost of bandwidth—depending on the proximity of the sensor to the Internet router. The cost of battery life is also significant as a large amount of power is consumed in transmitting and waking up the connectivity module.
  • Local or Cloud: IoT applications, due to the volume and cost, need to have the ability to process data “In-Network” or “In-Cloud”. “In-Network” or local processing of data becomes very important if cost of connectivity or speed of action is paramount. At the same time, a local processing engine needs to work as an extension of a cloud engine.
  • Control: There needs to be executable actions of the data being analyzed. If for example, Turbine data pattern shows that it is going to malfunction in a few hours, an “Action” of shutting down the turbine or immediately sending a service technician is required.

K2’s Action Analytics ™

“IoT” and “Predictive Analytics” have become the buzz words in the industry and every company is providing a wide range of solutions that predict the problem of “How” and “When”. What’s missing in the picture is “Why” it happened and “What Actions” needs to be taken when it happens. K2 is climbing the IoT mountain with its differentiating platform that assimilates data and business rules to create IoT Analytics application that will Analyze, Predict and Prescribe Action plans.


How it works

  • Looks at past performance and understands that performance by mining historical data in order to look for reasons behind past success or failure.
  • Provides Action analytics and answers the question –What will happen? This is when historical performance data is combined with rules and machine-learning algorithms, to determine the probable future outcome of an event or the likelihood of a situation occurring.
  • Goes beyond predicting future outcomes by also suggesting machine-learned actions to benefit from the predictions; shows the implications of each decision by enabling the application to provide the business rules that will allow determination of actions.



Use Cases



Remote diagnostics, Embedded software, Bluetooth LE, Device management, Predictive maintenance, UI portal.



Weather forecasts, On-the-ground sensor data measuring moisture, Soil pH, and temperature.



Smart sensors to collect data e.g., traffic conditions, bus location, speed of transit, engine health & video feed.



Smart Thermostat, Connected Door locks, Motion sensors, Video surveillance, Smart watering.



Connected training equipment, Personalized training sessions, User profile management, Remote access.



Event engine, Data analytics, Cloud platform, User preferences, Customer database.