Companies in the process industry often tell us they’re sitting on a pile of data. But as we dig deeper, we often find that the data is spread in different and isolated systems, on paper, in Excel files, access databases, and so on.
As a result, the process data is only valuable in the case of a customer complaint or an emergency, when a human being will look into all these data sources to find out what went wrong – and hopefully find the root cause.
Data Historian as an Insurance Contract in the Process Industry
The data historian’s primary role is to store time-series data from a variety of field devices such as sensors, valves, and so on. Once stored, process engineers can use it to analyze process performance by retrieving the data in a trending tool.
This does not differ at all from how process historians were used decades ago when these systems started making their appearance in the industry. From our experience, this is where most industrial companies’ data journey begins: process engineers having a single source of truth of process data that is easily retrievable in case they need to find the root cause of some problem. Therefore, the process historian bears similarities with an insurance policy in two ways.
The first is that, in an ideal world, you wouldn’t need it. Your process would be running perfectly all the time. The reality, however, is different. In that case, you want accurate and reliable process data to be available, just like your insurance company is covering you when something went wrong.
Second, in some industries, it’s mandatory to have on record what happened in the process, such as in food or pharma for traceability reasons. Just as some insurances such as liability insurance or occupational accident insurance are mandatory in lots of countries.
However, we have yet to encounter the first person that happily pays an insurance fee. So, why not turn that data into an asset?
Impact of Automated Data Collection on Organizational Transformation
Some organizations manage to create new value from the data they collect, often in ways, they couldn’t envision at first. People with different responsibilities within an organization become curious and start asking more questions. As a result, they start expecting more from the process data they collect.
Even though this transformation happens gradually, it can be sped up drastically with the help of a catalyst. This catalyst is making the data available for everyone in the organization, rather than just the process engineers or the IT/OT people that know what a tag is and at what memory location they need to grab it in the PLC.
Real-time data insights help people collaborate. Making industrial data available to anyone in the organization has become trivial with modern, open-source IT technologies such as the visualization tool Grafana and time series database InfluxDB.
These tools are so easy to use that anyone can get insights out of them pretty easily with hardly any training. And this can be done on the live historian, not some month-old backup in an Excel file. The performance is there and we have many cases to prove it.
Data For Everyone
Making data available to everyone is sometimes met with resistance when we talk to managers. There seems to be a fear of people asking too many questions or, more recently, a company that feared installing a ‘Big Brother’ culture because anyone would be able to view process data from everywhere.
In my experience, the opposite is true. Overall, people are involved to care about the companies they work for and will use the process data to improve and make their and their colleagues’ jobs easier. One example is the biomass plant A&S Energie, where all operators have access to the process data through their personal Grafana login.
Some operators have created dashboards that even go beyond our imagination. What about a dashboard that helps you start the steam turbine of the power plant? No problem! This operator’s dashboard has the turbine startup flowchart on the left and all the corresponding live process values from the historian on the right.
Industrial IoT as a Company Asset
Industrial companies who manage to see the historian as more than just insurance but as a tool to enable people and the organization as a whole are the ones that really reap the benefits of collecting and visualizing process data. Why? People start making data-driven decisions and embedding this way of working in the organization.
We have talked with production managers in the process industry who admit that a lot of the knowledge sits with the operators, explaining “that’s how they have always done it”, and thus admitting that sometimes tuning the process is more art and experience than science and facts.
By enabling people to collect and visualize any industrial data, it gradually becomes an asset in addition to an insurance policy. And as great assets do, they increase in value over time as more people and roles embrace it.
Data Historians Fuel Business Model Innovation
According to HBR, business model innovation is about delivering existing products that are produced by existing technologies to existing markets. And, they pose, because the underlying mechanics are often invisible, it is often hard to copy.
As an example, for some water business models, companies invoice the customer for the volume of treated process water, not by selling equipment. To do so, it needs reliable data collection to calculate the volume of water delivered. And they can even take it further, by integrating the historian data with the ERP system to allow for automatic monthly invoicing.
This is a great example of business model innovation in the process industry: the product is identical (clean water), the technologies exist (historian) and the market is the same as well.
Finally, it is not trivial to copy this new Water-as-a-Service business model because of all the pieces involved: process data, financing, streamlining maintenance, monitoring operations, etc. From practice, we see that it takes time for an organization to make the shift to properly embracing the potential of process data for more than its process engineering tasks. A bit like a flywheel that’s hard to spin initially but yields great momentum for the organization once it’s spinning.
Businesses in the process industry should start now. Identify potential improvements that go beyond the data insurance level but don’t take it too far in the beginning. Start small, design your data architecture, and get the naming right. Work iteratively with people in the organization to work out what strategic value can be extracted from the data.