Karthikeyan N, SVP & Head of Integrated Engineering Solutions, Tech Mahindra
The 21st century has been witness to significant strides in technology. One of the direct consequences of improvements in technology has been the growth of data. For instance, the volume of data that we have today is just a fraction of the volume we will have in 2020. And, over a third of all that data will be either on the cloud or transferred through it. We consume over 29,000 petabytes of video a year, which will bloat to over 3 times as much in 2018. Compare this to a total volume of just 75 petabytes at the turn of the century and we know we are in the middle of a data explosion. Can you believe that more than 75 percent of the data is generated over the last 3 years.
“The ability to generate actionable insight will help optimize systems that enable hundreds of applications and use cases”
A big chunk of all the new data will eventually come from machines that are connected to the internet, either directly or indirectly. As of 2011, there were an estimated 9 billion such connected devices, a little more than the globe’s population. Very soon, we will have over 24 billion such devices amongst us, each contributing data. Every home could eventually have over 25 devices that are connected to each other and the internet.
This unprecedented explosion in data is not just limited to volume. We are seeing new kinds of data that are complex, diverse and linked in ways that we do not fully comprehend. While the internet is the primary driver for this explosion, the Internet of Things will account for a large portion of the data spurt. A lot of the data of the future will actually be sensor generated and through machine-machine (m2m) communication. We must devise ways to use the huge amount of data because it offers us opportunities to innovate and disrupt. The possibilities created by the co-existence and mingling of machine data and human-generated data are too immense to ignore.
Existing technologies and paradigms for data management will be stretched and, ultimately, be rendered redundant. We need new ways to work with and understand all the new data. While evolutions in in-memory storage, advancements in sensor technologies, analytics platforms and Big Data technologies have enabled capabilities in gathering and processing large streams of data, deriving insight and intelligence from the data is the key. And, data to insight is a huge leap. The ability to generate actionable insight will help us predict outcomes and optimize systems than can enable hundreds of applications and use cases, including:
• Real-time systems
• Affordable, connected services
• Design of Predictive and cognitive systems
• Security and Privacy
• Next generation User Experience
Engineering is one space where data driven insight collected from sensors can deliver tangible value and improvements in productivity and efficiency. Engineering Analytics (effective way of analyzing it through physics of failure and data analysis) is what will make it happen. Tools and techniques that can help us visualize seemingly complex relationships or glean latent insight from diverse data are pushing Engineering Analytics to the forefront of innovation and progress everywhere. Analytics and algorithms that harness the raw power of large data sets will redefine how we do things in the future and help us make the leap. It will present companies the ability to adapt and tailor solutions based on unique customer experiences by analyzing real time data to measure even intangible factors such as sentiment and impact. Considering the immense possibilities, Engineering Analytics will form the basis for and drive the shift towards next generation connected solutions and services.
Nowhere is Engineering Analytics more relevant than in modern cars. The automotive/ aerospace industries are poised to become the world’s second largest data producer. It is estimated that sensors in modern cars and aircrafts generate over a gigabyte of data every hour. Combine this with collected and available data about driver preferences, usual travel times, fuel economy, average speeds, location data, data on weather/road and air traffic conditions. We have something truly meaningful. For example, an insurance company can use this data to gauge driver behavior. Insurance payments could be potentially smaller for drivers who drive very little, during off peak hours, in low traffic zones. Every additional data set that becomes available creates potential for a new business or a new way of doing things.
Engineering Analytics can also be applied in modern factories and industries. Factories and production environments are equipment intensive.Connecting all the equipment to a platform and harmonizing communication protocols and data can improve operational efficiency. Potential industrial applications include the ability to monitor and predict potential failures and breakdowns of critical equipment through sensor enablement and predictive analytics. The smart, connected factories of the future will have never-seen-before productivity and efficiency.
With machine data and human data co-existing seamlessly, we must develop analytics frameworks, algorithms and techniques to transform the data into intelligence. The strides we make in developing powerful analytics could well determine the ubiquity of IoT based systems and solutions. We expect the industries and companies who tend to master the simple use cases to dominate the market over the next decade. IoT is leading to exponential innovation by building products as platforms which engage with consumers through the life cycle beyond its core purpose and enhancing the value delivered to the consumers.