Nowadays data is everywhere. It is the hot topic of the moment, for businesses and in the general public. There are heated debates in politics about “Data”, data wars, and even a new science entirely dedicated to data!
When it comes to data, you might be wondering, should you build an enterprise architecture with data as your starting point (data-centric)? Or build a data structure around the existing landscape (data-driven)? What about data lakes? Do we actually have any data?
Well, tackling data is highly dependent on the specific configuration of your organization history, business, and IT landscape.
At Cuurios we believe that Data should be CENTRAL to your organization. Gathering, managing, and acting on your data is your core business. But there are no one-size-fit-all solutions. Data is a mindset, the most important is to just do it, however small the first steps might be.
In this blog post I am giving some insights on how we see Data at Cuurios and sketching the first steps towards data proficiency!
It sometimes feels like data is something new, something very hot, the core of the 21st century technological battlefield. But data has been there for a long time! Measuring, gathering and analyzing data is at the core of the scientific and industrial revolution. Without Galileo gathering data on the moons of Jupiter with his telescope, there would have been no proof that Copernicus was right, and we might still be thinking that the earth is at the center of the universe.
Data is not a thing, a disincarnated entity that exists for and of itself. Data is grounded in reality, it is information that represents assets, people, events in the real world. Data is what makes large-scale organizations possible. Without data, how would you know what the state of you inventory is without having to recount every time? What the state of a critical asset is without having to look at it?
The first form of writing ever discovered, the Sumerian tablets , were accounting records of production and exchanges of goods, i.e. data.
What has changed to make data the focus of a new gold rush?
- Storage capacity has increased exponentially. When a Sumerian scribe needed hours to imprint a clay table, we can now record terabytes of data for very little cost.
- The internet (as in the complete networking infrastructure). No need to have people do the measurements and record the data themselves. Everything is be automated.
- Advanced in computer power has made the application of advanced AI algorithms cheap and rewarding.
This list describes techniques to store, manipulate, and analyze data.
What it does not describe, is a change in the nature of data. People tend to concentrate on the new hype, assimilate data with data science, and equate analysis with machine learning. This is a very narrow view of what data is and limits its usefulness to a few very advanced use cases.
Because first and foremost, data is information, information about your business, its customers, its assets, its financial state. It represents the tangible and is often the only thing you have to steer large complex organizations.
We think that data should play a core role in every organization, be CENTRAL to decision making and action taking. Without data, any decision taken is an educated guess.
THERE IS MORE DATA AROUND THAN YOU THINK
In practice, we often encounter organizations that claim not to have any data. Because they don’t have a data lake.
The first thing they have on their data roadmap then, is to create one. But really, a data lake is just a big database. It won’t tell you what to do with your data. In our experience, many organizations, after having spent an incredible amount of time and budget on creating a data lake, are stuck. They don’t know what to do with it.
Because your data is about your business, not technology, not algorithms.
What we usually see is that organizations already have data, very often plenty of data, scattered around, in custom made applications, asset databases, excel sheets. Because you can’t function without data.
What they lack is an approach, a concrete process to manage data, to embed it in its day to day operations. The data processing, the algorithms, should come in support of operations.
Making sure data is part of your operations day to day business, that is being data-centric.
In order to do that, you need to reverse the data analysis process, look at your data from a business perspective:
- What are my most important use cases, processes, assets?
- Which data do I have about them? Where can it be found, in which format? Do I need more of it?
- How can we automate this specific use case? Which algorithm can be used?
At Cuurios, we have extensive experience working in the industrial sector. In most Industrial settings, processes will already be described. They will be backed by data for real-time monitoring, stored in a historian. Optimization and analysis algorithms are known.
The actual running of the algorithms, the analysis of the data and the generation of advices is traditionally a step performed by engineers, in Excel, Matlab or others. Tools great for exploration and scientific inquiry, but not made for automation.
For most companies, there is tremendous value to be added by connecting data sources to each other and automating their analysis. Actions can be defined and set-out quicker, with a better response to issues and a higher efficiency.