The Internet of Things and Industry 4.0 continue to dominate industry media, with ongoing attention to how these technology initiatives will increase productivity and profitability. Since then, however, the narrative has evolved. From the sensor to the cloud model to the advanced computing model as an essential component for real-time analytics and as an intermediate data collection / storage point. Firmly believing in this evolution, it is always good to be able to hear firsthand what is really going on inside industrial companies. The Internet of Manufacturing (IoM) conference in Chicago has proven to be a good place to interact with both suppliers and IIoT end users.
Let's divide the end users into two separate groups; those already on the road to IIoT; and those in various stages of assessment and implementation. The most discussed projects in the media:
preventive maintenance
digital twinning
machine learning applications
Enterprise Integration
Everyone uses data from different parts of the organization to streamline operations.
The common thread here is that successful implementations depend on the convergence of OT and IT. Without the successful collaboration of these two organizations, it will be very difficult to successfully implement IIoT projects. This is because the data is taken in the enterprise (and for some applications elsewhere in the company) and sent to repositories inside and outside the OT environment. There, various forms of analysis are carried out to achieve the desired goal.
Taking appropriate action
New applications and data transfers outside the factory are often a common concern for automation engineers. They disrupt existing control and cybersecurity processes. The ability to continually add new applications using technologies such as virtualization and cybersecurity layers is often the domain of IT professionals. However, as advanced computing penetrates deeper into the production environment and moves away from IT and even OT, it becomes more important that computing platforms natively support virtualization technologies. Everything except self-protection and ease of maintenance. The only thing we can say for sure is that any form of analytics does not perform well on data loss!
While IIoT is often associated with machine learning and artificial intelligence technologies, companies are finding that there is a lot that can be done with existing data. And all this without the need to immediately invest in these areas. Depending on the age of the existing control systems, it may be necessary to supplement the assets with additional sensors. Careful consideration should be given to retrieving (additional) data for analysis and storing it locally.
One of the companies attending the Internet of Manufacturing Conference was struggling with just this problem. Their modern PLCs lacked the computing power to retrieve and process data along with the execution of control loops. With minor infrastructure changes, a high performance computing system is ideally suited to the requirements of this company.
In some situations, benefits can be obtained simply by linking system processes together. This is achieved by replacing manual data collection and recording with electronic systems. Result?
It eliminates errors and provides almost instant feedback with relatively simple graphical correlation tools. This is an area where reliable high performance computing can play an important role.
Eliminate serious problems
Most of the Internet of Manufacturing attendees were in some combination in the evaluation or planning phase. In some cases they had a clear idea of what they wanted to achieve, in others they were trying to figure out where to start. Some of them have implemented pilot projects and experienced difficulties with large-scale deployments.
Several people expressed the following three main concerns:
Shop delay
cloud implementation
Cost associated with scalable cloud deployments
And while many people have conceptually realized the benefits that using the cloud can bring to an IIoT project, the requirements that will make this type of implementation ineffective are as follows:
Real-time feedback for reactive process control
Quality adjustment
Prevent imminent machine failure
For those concerned about rising costs, LTE mobile connection and cloud computing costs have dropped significantly and are still often dependent on the number of users. In addition, they fluctuate and grow from month to month as the deployment scales and the volume of data increases.
newyorkersblog cosmopolitansblog realsimpleblog nextwebblog theinformativeblog
