Picture yourself tasked with creating an essential report. You start by collecting information from various sources, initially dumping everything into one document. This document serves as your melting pot, where you begin to give structure to chaos—sorting, categorising, and highlighting what’s crucial. For this specific report, you only take from it what you need. For a future one, you’ll go into your well-organised document and pick other sections to use. 

That, my friend, is why a platform that gathers and structures your production data, and then uses it for different functionalities, is key for your productivity. 

I’m going to use this analogy throughout the article, I think it’s much easier to understand. We’ve all been through these situations, and the machines on the plant floor are weirdly similar. 

The downside of having no single source of truth

The scenarios I’m about to describe come from our experience as Premier Ignition Integrators, and the hurdles industrial companies were experiencing before using a platform to collect and store data for further analysis.

Wasted resources and poor decision-making

In life: You run from one person to another

Firstly, because you don’t have a way to gather information automatically from everyone in one place (e.g. a Google Doc where everyone can enter their info), you physically go to each person individually. 

You ask each person, in hierarchical order only, for information, and you have a separate piece of paper for each. Naturally, you’ll end up with lots of messy pieces of paper, instead of one digital source of information, where everyone pitches in AND from which everyone gets the information they need. 

To put information in one section of the report, you have to go through each piece of paper individually and try to find the information you need. You’re wasting a lot of time and effort on that.

Secondly, there’s an important dimension: time. The process above, manual and tedious, implies that data doesn’t come in from different sources at the same time. Once you get to the next person, the data from the first one may have changed. Your data sources won’t be aligned from a time perspective, thus putting you at risk. 

On the plant floor: Limited processes and plenty of risks

Limited processes

On the left side, you have the classical way of doing things: the automation stack. Stack them on top of each other.

The problem with this model is that each layer in this stacked model is a “peer-2-peer” communication. For example, the SCADA system communicates directly with the PLC, but there is no easy way to communicate directly with the financial system. If the Finance department wants information from the plant floor, they can’t get it easily. 

The consequence is that there is NO EASY way of flowing data across the stack. You can get data out from each layer of the stack, but no automated or effective way to get all data out at the same time and use that data across the entire business. That can lead to all kinds of problems: miscommunication between different departments, poor or slow customer service, missed opportunities, or uninformed decisions.

On the right side of the above image – everything is connected to a platform.

On an abstract level, everything becomes a node in a system: the PLC is a node, as is the individual user the financial system, or the database. It’s a Hub & Spoke model, and the platform in the middle is Ignition.





Risks because of inaccurate and incomplete data

I mentioned above the dimension of time. If you don’t have real-time data coming in, you’re taking shots in the dark.

For example, a utility company needs real-time data about pumping stations, the weather (fx. heavy rain), the (over)flow of rivers etc. If you have that coming in in real-time automatically, you can answer questions like “why did we have a spill-over in that area in those hours?”. You compare data with the same time stamp from multiple data sources and you find your answer. If you have it all coming to a platform, that is. 

Inefficiency, errors, and tedious manual processes

In life: You need specialised translators for each person

To continue the analogy, imagine trying to compile a report and needing to consult various colleagues for information. However, each colleague speaks a different language, forcing you to hire a new translator each time to understand their responses. This constant need for translators makes gathering information not just slow and costly but also prone to errors in translation.

On your plant floor: Siloed data 

The scenario above mirrors what happens in an industrial setting. Each machine or system, much like each colleague, operates using its own “language” or protocol, complicating direct communication. 

Here’s where Ignition shines: it functions as a universal translator, enabling clear and seamless communication among all devices. With Ignition, not only is machine-generated data effortlessly translated into a common language, but machines can also communicate with each other. Ignition can also send commands to the machines in a language they understand. 

Machines and systems are arranged in layers, much like the floors in your office building. At the bottom, you have sensors and PLCs (like the ground floor speaking English). Then, as you move up, there are other systems like MES, ERP, and eventually the cloud (like the other floors speaking their own languages). Just like in your building, data needs to be translated as it moves from one layer to the next to reach the top or to come back down. This setup makes sharing information between layers complicated, time-consuming, and prone to errors. Adding a new machine (or “language”) means more translation work, making the system hard to scale and maintain.

Additionally, because each layer only directly communicates with the one immediately above or below it, data can’t move freely throughout the organisation. Some valuable information might never be used because it’s too difficult to access from other layers. Trying new things or making changes becomes a big challenge because you first need to navigate this complex translation process to get the data you need.

A practical example involving real-time data

If you sit in the control room and see a need for more power somewhere in the grid, you need to fill that void. 

That means real-time access to the assets: what do we produce right now, what is the potential, if any? Plus, it means real-time access to the market about prices and capacity. And, additionally, you might need real-time data about the transmission system to get the production from the assets to the grid. 

In total, a rudimentary problem, but a complex task involving several different systems in different “layers”. Here, again, the platform idea is valuable. 

In essence, siloed data from different sources that don’t share a language means wasted resources, a high possibility of error, and unnecessary manual work. Paired with the lack of real-time data coming in from different sources in a contextualised and meaningful way, it ultimately means you can’t use data to work smarter.

At a glance: why use Ignition as a platform?





A quick word about the future: Machine Learning needs a platform

Most of us are thinking about advanced machine-learning models for the future of our operations. 

We all understand by now that we need heaps of data. We understand that we need valid and contextualised data. We also understand that we need to collect truckloads of real-time data for a longer period of time to have enough historical data to allow for training the models. 

However, we might not completely understand what type of data would be useful or necessary for the model. Thus, by storing valid and contextualised data from multiple sources in one unified place via the platform, we have that data available when we decide to start the Machine Learning journey. It’s long-term thinking, and it matters. 

We hope the analogy was useful to you. The whole idea is to move away from manual, tedious, error-prone work to an automated and well-integrated way of doing it. We have all this tech these days – much of it is useless, but this type actually helps. If you’re going to use any tech to work smarter, this is it.