Is IoT the next frontier for data scientists?

By Geetika Goel, CTO, Jigsaw Academy, The Online School of Analytics

The Internet of Things (IoT), a popular buzzword today refers to an ecosystem that integrates people and devices (both mechanical and digital) by connecting them to each other. Equipped with sensors and unique identifiers, these devices are capable of transmitting data over a network without requiring human-to-human or human-to-computer interaction.

Geetika Goel, CTO, Jigsaw Academy, The Online School of Analytics

IoT data transfers are almost always voluminous. The data thus generated must be analyzed if any valuable insights are to be drawn from it. In order to be successfully utilized, such large quantities of data require extensive analysis, a complex task that data scientists are probably best equipped to undertake.

At this point, it is worth asking whether data scientists are better suited to jobs in the IoT sector than fellow scientists from allied fields Or is the data scientist just one of many experts required by IoT ecosystems to help them function optimally? Conversely, will IoT be the next frontier for all data scientists?

At Cypher 2016, a data analytics event that ran from September 15 to 17, industry professionals sought to answer these and other pressing questions. At a panel chaired by Geetika Goel, CTO of Jigsaw Academy, eminent panelists Srinivas Padmanabhuni, Co-founder at Tarah Technologies, Kavitha Mohammad, Director, Industrial Solutions Group at Intel Corporation, Raghavan Kirthivasan, Sr Manager at AIG Science and Sanjay Srivastava, Director – Analytics at a Fortune 500 company, addressed the following topic: “Is IoT the next frontier for data scientists?”

A few years ago newspapers spoke about drones flying around and how they had a plethora of applications, but it’s usage had to be scrutinized for security reasons and the entire drone market was regulated and was even banned in selected circumstances. Similarly, even before qualifying IoT as the next frontier, one has to start by answering the critical question as to whether the buzz around IoT is just another swank idea or will it change lives and is here to stay!

Examples from various fields suggest that IoT is not only useful to the end consumer with cool use-cases like a fit band, a drone flying around, etc. But the field finds its use in industry applications like Healthcare, manufacturing, automobile, Home Automation, construction, etc.

For example: The manufacturing industry stands to gain from IoT because the costs invested in these manufacturing units are huge, but when these expensive, huge units break down, not just money, but even valuable time and resources are lost because of unexpected accidents. However, IoT devices can help in predicting when this manufacturing unit needs servicing, when the unit will not work even after servicing, whether there are any impending breakdowns, when to decommission and install new machines, etc. These devices don’t just manage inventory, but can efficiently regulate an entire supply chain. IoTs thus have tremendous mainstream appeal, because various industries will be able to adopt, customize, and incorporate them into existing systems.

IoT’s applications therefore is so wide that it’s definitely going to become a must for almost every business function, and that’s why IoT is here to stay.

That brings us to the next question of why is it exciting for a data scientist and what’s different about analyzing IoT data? It is different because the data in IoT is coming from so many different devices, in different formats, sizes, frequencies and quantities. So data scientists need to tweak their algorithms to be able to manage this huge amount of heterogeneous data, and that makes it different, interesting, and exciting for them from analyzing regular data sets that is usually generated from current systems and processes.

While IoT has the potential to be enormously beneficial for various consumer and business applications, the technology is still at a nascent stage and will take years to mature. Scientists envision a sophisticated ecosystem that deploys an army of cloud application and mobility engineers, as well as embedded hardware technicians who are trained to setup and configure device level specifications. In addition to this, they would enable these devices to connect to the cloud, allowing data scientists to access and analyze data streams in real time, and enabling them to improve quality of the services provided by these IOT networks.

Work is constantly on towards reducing IOT costs and a big breakthrough has happened. Full duplex radio has become a reality as of sept 14th 2016 and that is set to reduce IOT communication costs significantly.

IoT as a learning field will mature in the coming years, when experts in disconnected fields will be able to understand and contribute to larger IOT solutions. As we see it, where analytics as field was 5 years back, IOT is today.