Can Connected Factories Really Attain ROI?
Can Connected Factories Really Attain ROI?
Let's take a look at a few ways factories are embracing IoT and what the early indicators of success — and failure — have been.
Join the DZone community and get the full member experience.Join For Free
Digi-Key Electronics’ Internet of Things (IoT) Resource Center Inspires the Future: Read More
The ubiquity of the Internet of Things means that we're seeing not only a growth in connected devices but a growth in the data that IoT devices collect. Over the next decade, we can expect to see an acceleration of data creation as everything becomes instrumented and interconnected. For example, it's predicted that a connected car will use 4TB of data per day with over 200 sensors per car, making it effectively a data center on wheels. This is far too much data to send to the cloud for processing and return to the car, particularly in car control instances that require real-time processing for instantaneous decision making.
An area of business where IoT is making its greatest success stories is in the Industrial landscape. A sensor embedded machine on the factory floor, for example, can produce data that enables product managers to gain real-time insights into the operation of the machine — predicting flaws in the machines before they occur enabling the use of preventative maintenance. Precision analytics can reduce downtime and operational costs, and increase operational efficiency leading to substantial savings in time and money.
Yet while there are significant success stories when it comes to data analytics saving money and time, the sheer quantities of data at play can be a pain point. I attended Bosch Connected World in Berlin where I met with Kobi Suissa, a pre-sales Engineer at Crate.io, who was part of a slew of startups pitching to VCs and journalists. The IoT Innovator Awards named Crate.io Best IoT Open Source Software in 2017, and they won the Disrupt Europe 2014 Battlefield. They created CrateDB, a distributed SQL database that makes it simple to store and analyze massive amounts of machine data in real-time including processing IIoT data in real-time to improve factory efficiency.
Kobi explained to me that: "An example of Crate's work in action is with manufacturing company ALPHLA who produce plastic packaging products for brands like Coca-Cola. As a factory, they integrate sensor data from 1500 production lines (1 million sensors, 900 different types) into a central mission control staffed by experts who monitor production via real-time dashboards. Those experts give equipment operation and repair direction to less experienced workers on the factory floors."
The Problem of Speed
ALPLA had tried SQL server to manage the sensor data. Querying steams of thousands of readings per second in real time and running complex machine learning analytics on terabytes of historical data was outside of the SQL Server's capacity. Charts in the mission control dashboard were taking 3 to 5 minutes to process which was prohibitively slow since each dashboard contained over a dozen charts.
According to Kobi, "CrateDB is purpose-built for IoT workloads with functionalities such as distributed processing, and dynamic schemas meaning it could run ALPLA's queries 25x faster than SQL server." This meant that ALPLA was able to lower personnel costs by requiring fewer experts; gain immediate insights instead of next day reporting, and achieve cross-plant analysis to get new insights."
Connecting factories can be an expensive enterprise, particularly those that are still embedded with legacy architecture and machines, or a variety of different operating systems across a miss match of IT and OT. The end to end analytics needs to create meaningful insights to offer an ROI. Kobi explained that the promises of smart factories had seduced many companies, yet the slow speeds made their experiences less than stellar. It'll be interesting to see how companies like Crate.io will respond to the promises that IIoT has promised (but not always delivered by competitors), but also whether they evolve into front-runners in an increasingly competitive space.
Opinions expressed by DZone contributors are their own.