In my last blog, I focused on setting expectations for situations where selecting the appropriate edge hardware for connectivity also meant you may have limitations on the volume and frequency of data you can collect. It is important to understand these limitations. This is a great segue into making sure you have the right data to power the right apps that deliver business value, because IIoT is nothing without a viable business outcome.
Once the expectations are met on connectivity, the next step is making sure you have the right sensors or instrumentation to feed data into the application you are using to solve your problem. The myth that you can just use the magic words “industrial IoT” in your planning to solve most process problems is not going to get you far. The reality is, you still need to understand the fundamentals of instrumentation to pick the right sensors, and you need to understand the problem you are trying to solve to ensure you select the right application all while staying aligned with process operations, maintenance and the IT teams. There are many moving parts!
It is true that IIoT connectivity can help overcome infrastructure issues and that you can now add wireless sensors to pick up just about any signal out there. Despite these new capabilities of edge connectivity, we do not want the connectivity part to mask any real issues. We need to make sure we are solving a real business problem and not just adding a cool factor. In some cases, visibility of the data is the business need on remote assets; however, in other cases this new ability to collect data from what was once a stranded asset is not enough to drive real justification for doing an IIoT project.
Think About the Goal You're Trying to Achieve
In most cases today where we see the use of IIoT applications, we see customers trying to solve complex problems that involve many data variables with intricate relationships that demand the use of big data tools and large amounts of compute power. This is where apps for AI, multivariate analysis, pattern recognition, event detection and machine learning all step in. It’s important to select the right application for the right problem and the right process.
The app you choose should be tailored to the unique needs of the asset and the process, continuous or batch. For example, if you are trying to solve a quality issue on an asset, you would most likely need a multivariate analysis tool, but if you want to implement predictive maintenance and early failure detection using dynamic models, you would need a machine learning app. Selecting the right app for the problem you are trying to solve is critical to drive business value.
This is still only one of many decisions you need to make in your IIoT deployment. Should you use a cloud-based app or an on-premise one? Does your solution require high-speed computing with low latency driving the app closer to the edge device? These decisions now need to factor in costs that affect your return on investment (ROI). Your solution may have cloud service charges, data transmission charges and labor costs that you need to consider for your Industrial IoT solution deployment.
Selecting the right connectivity, the right sensors and the right app are all important to driving the right business outcome, but managing the deployment costs is also a big factor in driving success. On top of all that, the choices for each part grow by the day. Let us know if we can help!
To learn how AspenTech can kick-start your IIoT initiative, check out the Aspen Connect family of solutions.
Leave A Comment