How to get your device data unified with ease

For companies across manufacturing industry, the process of data normalization is one of the key challenges when it comes to scaling Industrial Internet of Things (IIoT) solutions. Often it takes a lot of time to transform heterogeneous data from different kinds of devices into a unified format which is needed for deploying AI solutions. Omnio, a Copenhagen-based partner startup of 5-HT, has built a solution that speeds up the process of data normalization. “By reading the manuals of thousands of industrial devices, we have built a vast library of knowledge that helps you deliver your data ready-to-use to any platform”, says Mikkel Schönherr Thøgersen, co-founder and COO of Omnio. In this interview, he explains how his startup enables manufacturing companies to easily scale their IIoT solutions.

Mikkel Christian Sørensen and Mikkel Schönherr Thøgersen winning the SEED Discovery Day 2017

How did the idea for Omnio develop?

Back in 2009, I joined the Danish logistics company Maersk where I teamed up with Mikkel Christian Sørensen who later became my co-founder. He had the idea to connect all of Maersk’s refrigerated containers to the internet in order to determine their exact position and to enable predictive maintenance. In 2009, the Industrial Internet of Things (IIoT) was not very well developed, so we had to build all the different layers and components by ourselves. In this way, Maersk’s “Remote Container Management” became one of the world’s largest IIoT solutions at that time. Today, it still connects all of Maersk’s refrigerated containers around the world. We were one of the first teams to scale an IIoT solution to a global solution and to come across all the challenges you face in the scaling process.

What are the challenges you experienced when scaling your solution?

Because Maersk used many different types of containers from different manufacturers, one of our biggest challenges was to normalize the data so it could be used by an application. This process took us about a year. Soon we learned that we were not the only ones who had trouble getting data normalized out of devices – this was a problem that occurred across various industries. For example, different devices may measure the same thing while using different units. If you want to implement a monitoring, power management or process optimization solution, in the beginning you often need to spend a lot of time on normalizing all the data from your different devices. Once we saw that this was a general problem, we decided to solve it not just for us, but for everybody else. This is why we founded Omnio in early 2017. After winning a number of pitch competitions, we were invited to join the Techstars New York City Accelerator in summer 2017 where we got introduced to lots of interesting companies like SAP. Today, there are around 20 people working for our startup in Copenhagen.

Who are your customers?

We serve small, medium-sized and big companies, mainly from European countries, but also worldwide. Our customers are companies across manufacturing industry, e.g. chemical companies or utility providers. They typically have a lot of industrial equipment, which is why it makes sense for them to optimize their assets, for example with remote monitoring, predictive maintenance or energy management solutions. In many cases, however, our customers have made the experience that they tried to implement an IIoT solution but got stuck in pilot projects because they found it very hard to scale. For example, if a customer wants to implement an energy management system which promises to save him 15 % of power, the solution must not become too expensive to roll out – or otherwise it will take away the benefits. With Omnio, we are able to drastically speed up the implementation of IIoT solutions. The normalization of data that took us a year at Maersk in the beginning is now something we could do within a few weeks. As we reduce the costs of data normalization, we make IIoT solutions affordable and profitable for the industry.

How does Omnio manage to accelerate the process of data normalization?

By reading the manuals of thousands of industrial devices, such as Frequency Converters, Meters or PLC’s, we have identified all the relevant information and built a vast library of knowledge about the individual communication profiles of the different devices. In these profiles, we map various kinds of device data including the formats and units in which the data is delivered and requested. This library is the foundation of our software solution that collects raw data and delivers it unified and ready-to-use on any platform, application or datalake. We have built integrations for ThingWorx and Azure IoT, but we can also help our customers to integrate our product into any cloud using standardized APIs.

How does the process work for your customers?

First, we have a technical meeting in order to understand our customers’ solutions and to introduce our product. The software is very easy to use: our customer gets an interface where he can simply select the devices he would like to connect. This takes maybe 30 seconds per device. Using Omnio as a middleware in their own solutions enables our customers to unify the data from their industrial devices within a few days, compared to the months or even years that are often needed for data normalization projects today.

What are your next plans for the further development of Omnio?

In terms of technology development, we are currently working on some interesting new features. At the moment, the user needs to tell the solution what kind of devices he wants to connect, but in the future we want our solution to automatically detect and connect the devices. In terms of business development, we want to mature our relationship with a number of Fortune 500 companies we are already working with. Furthermore, we aim to expand globally, with a special focus on startups as our customers.

How can 5-HT help you in this regard?

Up to now, we have primarily worked together with big and medium-sized companies, but in the future we would also like to address small companies and startups. Because they have relatively few resources, they are even more affected by the challenges of data normalization than bigger companies. This is why we would like to help startups and small companies to offer a better service to their customers. Since 5-HT has already grown a huge network of startups, we hope to be able to team up with some of them. As there are lots of startups working on IIoT solutions right now, we want to go out and help them to scale their solutions for their customers. Furthermore, we also hope to get into contact with large and medium-sized corporates from 5-HT’s network and support them in achieving their goals with IIoT.