So you have finally decided to start your digital transformation. You have seen the benefits around digitizing your operations. By connecting your disparate systems, you can start seeing what is actually happening in your plant and become more efficient. And not just in one plant, but in multiple plants.
You know you can reduce energy, waste, and save time. Just think about what you could do if you could collect and analyze data as well as command and control the plants.
You know that if you do not take this step, you are going to be left behind and that could be at your own peril.
Where do you start?
I have been working with manufacturers lately on how to architect the best system in a manufacturing plant. They want to get data from all their plants, but the plants have different systems and are situated all over the world. The discussion starts with what kind of data wants to be observed, but it quickly moves to what needs to get into the architecture of the system and what makes sense based on security, availability and functionality.
We have implemented on-premises or local installations for many years. That is the tradition. But is that the best architecture?
The cloud is a clear alternative. But what about security and availability?
Let’s take a look at the pros and cons of each type.
We are all comfortable with this type of architecture. That is how manufacturers have implemented systems for decades. The systems are secure (or thought to be) and available.
The pros include security, network availability, and independence. You wouldn’t be dependent on outside networks and equipment.
The cons include ending up with islands of information, higher maintenance costs, blocked access to information when outside of the network, and restricted expandability.
Cloud computing is not new. Most of us are using cloud computing now with email and storage.
There is convenience to cloud computing, and it is very cost effective.
The pros include lower IT costs; easy expansion of cores, memory and drive space; the availability of pay-as-you-go services; lower maintenance requirements; wide availability (anywhere, as long as the network is available); the possibility of connecting multiple locations; and the availability of analytics tools (business intelligence, predictive maintenance, machine learning and artificial intelligence, for example).
The cons include concerns over data security and hacking, and the possibility that lack of network availability may stop production.
Instead of seeing this as case of either/or, let’s look at a hybrid model.
The command-and-control function could be located on-premises. This could still be accessed remotely if the application requires this. That way, any network loss to the cloud would not shut down the plant. Then the vast amount of data that you are collecting can be stored in the cloud to relieve the burden on IT and equipment. The data is then accessible remotely and you have access to all the tools provided by cloud services. The expandability in cloud computing allows you to start small and expand without having to bring in new equipment.
Manufacturers should consider the hybrid model to get the best of both worlds by using cloud and on-premises models. This provides the flexibility and security that manufacturers need to implement their digital transformation.
Today’s networks, software, and devices allow for very flexible architectures. Data, analytics and visualization can be stored locally and remotely. Multiple plants can be securely connected, and data aggregated.
Every plant is different, so you need to find the right balance of local and cloud platforms. When you are choosing a platform to collect and visualize data, make sure the system has the flexibility to meet your particular requirements.