According to Goldman Sachs, the Internet of Things (IoT) is going to be biggest wave triggered by the invention of the internet. The 1990s’ fixed Internet wave connected nearly a 1 billion users, while the 2000s’ mobile wave connected another 2 billion+ users. However IoT has the potential to connect 10X as many (28 billion according to some estimates) users to the Internet by 2020 – where these “users” are inanimate objects ranging from bracelets to cars. Adding data science to the mix, IoT and analytics can exponentially grow this trend.

However, there are several major challenges to overcome before we can realize the potential of this exponential connectedness. 

  • Improvements in sensing technology
  • Improvements in battery or power technology (scaling down, instead of up)
  • Increase in computing power
  • Improved data communication bandwidth
  • Ingesting massive amounts of data within a short time span and finally, the most critical 
  • Analyzing and making sense of this data

There are five different application areas from today’s vantage point which can leverage IoT and analytics and are ready for early adoption of this third wave: WearablesAutomobiles, Homes, Cities and Industries. Among these early adopters, industrial applications are the ones which can realize immediate economic benefits in terms of Return on Investment.

Consider the example of turbo machines or turbines. When the critical sections of such machines are connected, they can generate a huge amount of data in a fraction of a second. What does this additional data tell us, that we did not know before? The objective here is to do two things: first use the data to establish what could be considered as baseline or normal operating conditions for the equipment. Second, against the backdrop of this baseline, monitor trends and patterns in the data signals to predict or forecast potential faults or breakdowns. Forced downtime or even planned downtime for critical equipment such as these turbines can cost millions of dollars. Using predictive analytics, we can get a leg up on preventive maintenance.

Current maintenance practices rely much on human experience, judgment and intuition. Experienced human operators (for example, pilots) may have an exceptional ability to control and make right decisions at during specific situations. But as the number of parameters to keep track of increases, it quickly overshoots human abilities. In those situations we need software / algorithms to take the decision or at least support to take appropriate decisions. All these decisions could be programmed by a controller, based on real time measurement of parameters such as Torque, Temperature, Thrust load etc for monitoring the condition of a machine. Recent development in the Industrial Internet have provided an opportunity to collect real time data from these turbo machines which operate under harsh environmental conditions. It is clear that success of such programs depends on the following simple logic:

  1. Accurately measure and transmit the information from sensors in real time
  2. Process, Store and Analyze time series information received from sensors
  3. Take appropriate decision based on the analysis and convey the decision to the machine operator or controller to take further action.

Each of the above factors will trigger enormous developments in the associated technologies and businesses, ranging from IoT and analytics and big data, to visualization to mobile. These are indeed exciting times for the world of connected machines, the internet and the economy in general.

(Written with support from Vaibhav Waghmare).

Originally posted on Thu, Dec 11, 2014 @ 08:00 AM

Tags:

No responses yet

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

simafore.ai