The Philips Hue, a connected light bulb whose color can be changed via a wireless mobile app. The Nest Learning Thermostat, a device that learns a user’s domestic schedule and programs itself to lower heating and cooling bills and which can also be controlled via mobile app.
These two seemingly different deployments in fact have one factor in common: the cloud. The former utilizes Google’s app engine on the backend while Nest works via a cloud-based analytics engine, which sieves through large amounts of data to determine a user’s ideal home temperature.
The cloud plays a crucial role in driving the Internet of Things by taking the complexity out of cloud application architectures, shared the technical lead for Google’s cloud platform in Asia Karthikeyan Rajesekharan.
The true value of the Internet of Things can be derived with cloud-based analytics, which provides the ideal combination of Big Data and Real Time information, said Karthikeyan, adding the cloud will prove still more important as more Internet of Things projects and devices are rolled out.
Internet of Things deployments are comprised of four layers – the sensing layer, connectivity layer, services layer and application layer. Analytics belongs in the application layer.
Analytics deployments can be classified into two categories, Karthikeyan shared. The first concerns large amounts of data and simple analytics while the second deals with huge amounts of data and more complex analytics. The cloud plays a crucial enabling role for analytics deployments in variety of ways.
Apart from the ability to run analytics tools in the cloud without having to worry about building such tools, the cloud also enables the running of data driven experiments without storage capex concerns.
“High level analytics deployments will likely run into volume and velocity challenges and that’s where the power of the cloud platform comes into play,” said Karthikeyan. “If you move your architecture to the cloud, you are no longer constrained by how to store that info and any associated complexities.”
The cloud also provides the ability for the easy tapping of data from different sources. For example, sensory data could be analyzed in the cloud alongside data from other sources. “You could blend and mix in many different mode for analysis and it makes things easier when you don’t have to worry about capex,” said Karthikeyan.
One such example is that of Google’s sensing lab project, where distributed sensors were placed across a facility measuring various aspects such as pressure, ambient temperature, radio frequency and noise levels. Data from the sensors was fed into Google’s cloud computing infrastructure and analyzed using Google tools.
“Data can be written straight into the data storage layer, with the analytics layer written on top.”
Anyone considering an Internet of Things deployment should take advantage of whatever cloud infrastructure providers are out there, said Karthikeyan. “Think about what the cloud brings to the mix and particularly around how it reduces the take it takes to bring your Internet of Things application to the forefront.”