Follow Datanami:
June 8, 2018

Harvesting IoT Insights, from Edge to Core


The rapid proliferation of Internet-connected devices is fueling the on-going data explosion. The IoT boom is also exciting company leaders, who anticipate new business models generating trillions of dollars. However, most of the IoT data will not be used, which some experts say points to a need for heightened business clarity around the technology and tighter integration of end-to-end IoT solutions.

From predictive maintenance on the factory floor and oil well monitoring, to tracking of chronically ill patients and self-driving cars, the potential business benefits of IoT, if fully realized, are truly staggering. “If policy makers and businesses get it right,” a 2015 McKinsey&Company report states, “linking the physical and digital worlds could generate up to $11.1 trillion a year in economic value by 2025.”

And while there’s certainly lot of hype around IoT – the hype-o-meter for IoT is arguably on par with excitement surrounding big data, artificial intelligence, cloud computing, and blockchain — the smart folks at McKinsey say that IoT’s benefits may actually be under forecasted at the moment.

(Source: McKinsey)

“Our central finding is that the hype may actually understate the full potential,” the McKinsey authors write, “but that capturing it will require an understanding of where real value can be created and a successful effort to address a set of systems issues, including interoperability.”

And therein lies the issue: While the value from IoT can be monstrous, getting the value out will be very hard. It’s a similar issue faced by organizations that jumped into the big data game with overly rosy expectations, only to find that building and running a big data system is a lot harder than it looks.

One firm that’s been in the IoT trenches and learned some of the tough lessons of success is Altizon. Since it was founded in 2013, the company has sold its Datonis Industrial IoT (IIoT) Platform to more than 130 customers across several industries, including automotive, tire, steel, fast-moving consumer goods, oil and gas, electricity, and chemical. The company, which has offices in Pune, India and Scotts Valley, California, has 70% of its customers in the manufacturing industry.

According to Altizon co-founder and CEO Vinay Nathan, one of the keys to success for Altizon and its customers is the end-to-end architectural vision reflected in the development of the Datonis platform. Instead of requiring customers to wire together different pieces to handle the various processes required in IoT – from ingesting the data at the edge to acting on insights, and everything in between – Altizon has built the system for customers and made it available as a shrink-wrapped solution that can run in the cloud or on-premise.

“With enough time money and resources, you can build anything you want,” Nathan tells Datanami. “I think the key challenges, of course, are the ROI and the timeframe. There is a lot more to it than just being able to slap together a few pieces of code.”

Altizon used a lot of open source componentry from the big data ecosystem to build Datonis, including software like Apache Kafka for moving data from the edge, Apache Hadoop for landing machine data, and Apache Spark for powering machine learning and streaming analytics, to name a few.

Gartner Magic Quadrant for IIoT

Customers could use the same open source parts to build their own IoT platform, but it wouldn’t be fast and it wouldn’t be easy. “We have a lot of engineering work to productionalize and make it highly available and entries ready,” Nathan said. “That’s non-trivial when you look at it in terms of being able to do that in-house.”

Gartner recently published its first Magic Quadrant for IIoT platforms. The analyst firms looked at a host of IIoT platforms (including Altizon’s Datonis) and rated their capabilities across several key requirements, including device management, data integration, data management, analytics, security, and application enablement. Nobody made it into the Leader’s Quadrant, which should be a sign for how far we have to go.

Altizon was in the middle of the pack among other Niche Players, including Atos, Oracle, Accenture, Software AG, IBM, QiO, and Flutura. In the Visionaries quadrant were Hitachi, SAP, and PTC, which was the furthest up and to the right (which perhaps is why its distributing copies of Gartner’s report).

Bringing all the pieces together to succeed with IoT isn’t easy, but the cost of doing nothing may be even higher, Gartner warns. “CIOs who do not act to investigate and to create digital business opportunities through the alignment and integration of IT, IoT and OT will cede growth in core markets and eliminate the possibility of extension into market adjacencies,” the company writes.

Altizon’s Nathan offers a few suggestions for companies that are building an IoT system or thinking about building one. “You should definitely start small,” he said. “This is a very iterative kind of process. You may try three projects and, if you expect success in two of those three, you may choose to scale the one with the most likelihood of success.”

Altizon’s Datonis architecture

It’s pretty much the same advice that companies were given regarding their first big data systems: smart small, get a “win” under your belt, and build out from there. “You need to take that kind of approach as opposed to having a three-year roadmap of exactly what you want to do, which is what a lot of manufacturing companies are more use to,” he said.

The cast of players for an IoT will also be similar, and will include data scientists and data engineers. But since the IoT system has the potential to touch other parts of the business – from consumer devices to oil tankers to legacy ERP systems — it will also need to include business experts.

Don’t ignore the need to bring a diverse array of voices to the IoT table, Nathan warns. “What people grapple with in terms of challenges are well understood,” he said. “This is a multi-disciplinary thing, so you need a cross-functional team to drive it. These are standard things that are expected and anticipated. But unless they’re in place, things don’t hold.”

Related Items:

Why Data Scientists Should Consider Adding ‘IoT Expert’ to Their List of Skills

Built to Last: Laying a Framework for IoT with Enterprise Architecture

Collecting and Managing IoT Data for Analytics