How can manufacturers overcome 4 of the top IIoT data challenges?
By 2025, analysts predict that Industrial Internet of Things (IIoT) will generate an economic value of $3.7 trillion, but not all manufacturers are ready to reap these benefits.
In order to leverage IIoT and smart machinery, manufacturers must obtain the key to opening these doors: improving the efficiency of their data collection, storage, and analysis from these devices.
Unlocking the benefits of IIoT may seem daunting but Ian Uriate, CEO of Timbergrove, a Liferay Service Provider Partner and expert in IIoT and manufacturing solutions, has provided these 4 best practices manufacturers should implement to overcome today’s data challenges.
4 Key Elements of IIoT Data Management
Sensing Data
Data can come from many sources, including third-party systems, equipment sensors, and human-initiated data like uploading an image from a mobile app. We’re going to focus on the data that starts at a sensor (i.e., a physical device that can translate a real-life condition like pressure or vibration into an electrical signal).
Sensors are the link between the digital and physical worlds of IIoT, and it’s here where human expertise is critical.
A sensor can’t determine if it has been correctly mounted and calibrated. Additionally, humans are required to ensure proper capture and analysis of the data. For example, machine learning can play a role in detecting that something is not working and, based on historical data and changes in patterns, provide an alert to a human who will check on what’s happened and take necessary action.
After taking measurements and recording data, sensors must then transmit the data to a central location known as “the beach.” Depending on the circumstances, this might be via a wired internet network, a control network bus like Profibus or Fieldbus, wireless protocols such a LoRa WAN or xbee, WiFi, a cellular network, or even a satellite connection.
With potentially hundreds of sensors taking thousands of measurements each second, data volume can quickly become an issue. Today, though it’s cheaper than ever to send data, sending gigabytes of data an hour, adds up. One way to make data transmission more efficient is through edge processing, that is, analyzing data at the sensor itself. The sensor then needs to only transmit the processed output data to the beach.
Storing Data
But data transmission is only part of the challenge.
Very few businesses have the capacity to store this much data on their own servers. That’s why on-demand cloud computing platforms like AWS and Azure are so popular in the IIoT programs.The “on-demand” component means that clients can easily expand or shrink their storage allotment so that they only have to pay for the space they use. The elasticity of the cloud allows for flexibility in cost and availability of resources. Changes in production and manufacturing can be managed fluidly without penalty for “buying” larger or smaller virtual spaces. The ultimate goal is to have a single platform to access all sensor data at once on a consistent and reliable basis.
Analyzing Data
In examining how this data will be used, Uriate describes that it's important to understand that the IIoT data is rarely analyzed by itself. Instead, data is pooled with other relevant datasets into an enormous ‘data lake.’” He continues, “For example, an engine manufacturer may have a data lake that includes sensor measurements from thousands of customer power plants and other engine-powered equipment over several years. Or, a manufacturer may want to create a data lake of sensor and other assembly line data from facilities around the world to compare, contrast and analyze operational efficiencies and need for equipment maintenance. By consolidating data, it’s possible to present, evaluate, and apply data trends more accurately.”
Many manufacturers use the IIoT data today for equipment monitoring and prediction. The IIoT data analysis can help predict the optimal maintenance interval for industrial equipment based on external factors such as operating profile and even ambient temperature. If data analysis of the IIoT can make such detailed predictions, one may wonder why we don’t use the IIoT to directly control equipment for optimum service life.
“Unfortunately, even the fastest connections are just too slow to effectively control machinery in real time, but I predict that the introduction of 5G and 6G networks will completely change what’s possible with the IIoT as real-time remote control becomes a reality,” says Uriarte.
Accessing and Storing Data
One of the biggest debates in the IIoT revolves around data access and ownership. Today, the question of who has access to the IIoT data depends a lot on who’s collecting the raw data and the intended use of that data.
For the management of any sort of industrial facility, it’s easy to see how a single system involves components and products from different original equipment manufacturers (OEMs) around the world. Yet, all of these products work together physically to form a system that works for the manufacturer. Right now, this is not the case with an IIoT project. The equipment manufacturers tend to protect information about their equipment and send data directly to their own data hub for analysis.
“While OEMs are happy to send their clients the results of their data analysis, clients need to push for more access to the raw performance data if they want to make better real-time decisions at the system level. Clients hold the power to access more of their equipment data; they just need to work together to agree on a standardized data consortium,” adds Uriarte.
Even within the same company, data silos that isolate data sets often result in inefficiency, poor decision-making, and a lot of frustration.
Fueling Up IIoT Benefits
Let’s see how these best practices have worked for a Timergrove customer in the oil and gas industry. This company was using an IIoT program to monitor and improve their drilling performance.
One key measurement they collected was mud pressure. However, when analyzing the data, there was no easy way to tell whether a change in mud pressure was due to a collapsing drill hole or a team of operators intentionally raising the drill.
“When it came to cleaning up the pressure data, it was impossible to tell which team provided different pressure measurements because each team recorded data in separate systems. We created a solution that enabled the client to capture time-stamped events and match these events with recorded historical data to “annotate” the data so that the context of the sensor reading was clear for more accurate and useful daily reports”, said Uriate.
In short, technology used with the IIoT is a powerful tool that can completely change how business leaders and operating managers make informed decisions. The key is to avoid data silos and instead promote data integration wherever possible. This will help make your IIoT projects all the more valuable.
How Timbergrove and Liferay Can Help
Take the next steps in developing a successful IIoT project by establishing clear goals for your organization. More than 25% of IIoT projects stall at Proof of Concept due to the lack of a good strategy.
To combat this, Timbergrove and Liferay are offering a free consultation on your next IIoT project. In this technical consultation, we will analyze your specific business needs, analyze your current road map, and assess where you are on implementing that road map.