Digitalization is the key upgrade that every industry needs at this era. The industrial revolution has evolved to the current state through various phases. These are the key phases of Industrial revolution and we call it with a version in the order it is listed below (Industry 4.0, Industry 5.0 etc.)

  1. Mechanization
  2. Electrification
  3. Automation
  4. Digitalization
  5. Personalization

Most of the manufacturing industries are in Phase 3 and some are in their journey from their 3rd phase to 4th phase. The key goal of all these transitions are to

  1. Optimize the operations. Improve monitoring and visibility of the progress at all levels
  2. Reduce downtime
  3. Improve the productivity
  4. Reduce the operating cost

Now there are a lot of products available in the market to solve the above problems. I will be focusing on the key points to be considered for Industry 4.0 and 5.0 related areas in this article.

The important features you need to assess before choosing any Industrial IIoT platform for the digital journey of your manufacturing enterprise are given below.

  • Data Acquisition capability – This should not limit to the most commonly available industrial protocols. There should be flexibility for adding any custom protocols and data collection mechanisms from various sources. Examples include image, video, flat file, pdf, web sockets etc.
  • Secure data transfer – Ability to securely transfer the data across various layers within an industry. The network layers are complicated at some industries. The platform should be capable of tunneling data through complex layers securely.
  • Data processing and transformations (Realtime & Batch) – Capability to apply transformation logic to the incoming data in a real-time mechanism. This is very critical as there are a lot of parameters to be derived from the raw parameters in real time to create the digital asset model. The platform should support the additional of any transformation logic as a plugin or a rule. It should never be restricted to the built-in set of rules / transformation logics.
  • Horizontally scalable architecture – The platform needs to be horizontally scalable. Some platforms work well in the initial implementation with a small number of use cases, but problems arises once we try to scale it. So the underlying technology needs to be thoroughly assessed before the choosing someone.
  • Visualization Capabilities – The platform should have some easy ways to visualize the data (raw & transformed). There should be some built-in capabilities and provision to extend it to external visualization and reporting tools.
  • Data Modelling capability – The platform should be capable of creating a digital copy of your manufacturing unit. So every equipment will be digitally modelled within the platform. The data model should support templating capabilities so that the same models can be reused across multiple sites, plants and lines. Support for multi-data model is a bonus point. I have not seen much platforms that supports multiple data contexts. The way a thing is looked at is different at different departments within an enterprise. So it is important to create separate contexts that is suitable for their view. This context needs to be maintained at all layers for the same persona / audience.
  • Data Accessibility – This is a very critical point. Some platform is very good at everything, but the data will be locked inside. This will lock you down and you will be tied to the platform vendor for every. They will rip you for any of your data related requirements. Always ensure the data model and the data is accessible in real time and batch for further use cases beyond the platform. In this case, you can build your own dashboards, applications etc. with your own tools without the support from the platform vendor. For example, if you have a requirement to build custom reports using a reporting tool like PowerBI, the platform should support the data sharing or 3rd party tool connectivity in an easy way.
  • Flexibility to deploy in On-Premise and Cloud environments – The platform should support deployment models which works in On-Premise and cloud environments. The platform should not be heavily dependent on specific cloud services. Some platforms provide software as a service offering where the platform is hosted somewhere in their cloud where you/your IT don’t have any role. This is not a scalable model. If you have plans to scale, never choose this model. There is a high chance that you will be locked for ever.
  • Scalability to build more use cases – The platform should be flexible enough to build and extend the capabilities to more use cases.
  • Capabilities at the Edge – Once we get into advanced level use cases, edge processing or advanced processing within the OT layer may arise as requirement. This may include closed loop use cases also. As closed loop use cases are highly critical compared to the data ops, the platform should have proven records in the area of closed loop use cases and advanced processing at the Edge.

I have highlighted the above points based on my personal experience and this is my personal view.

Feel free to comment below if you have any questions or feedback.