My holidays are over. After reading and cycling a lot, it is time to focus again on business and future. Those of you who have followed my blog the past year must have noticed that I have been talking on a regular base about business moving to a data-oriented approach instead of a document / file-based approach. I wrote an introduction to this topic at the beginning of this year: Did you notice PLM has been changing?
It is part of a bigger picture, which some people might call the Second Machine Age, Industry 4.0, The Third Wave or even more disturbing The onrushing wave.
This year I have had many discussions around this topic with companies acting in various industries; manufacturing, construction, oil & gas, nuclear and general EPC-driven companies. There was some commonality in all these discussions:
- PLUS: Everyone believes it is a beautiful story and it makes sense
- MINUS: Almost nobody wants to act upon it as it is an enormous business change and to change the way a company works you need C-level understanding
- PLUS: Everyone thinks the concept is clear to them
- MINUS: Few understand what it means to work data-oriented and what the impact on their business would be
Therefore, what I will try to do in the upcoming blog posts (two-three-four ??) is to address the two negative observations and how to make them more precise.
What is data / information / knowledge?
Data for me is a collection of small artifacts (numbers, characters, lines, sound bits, …) which have no meaning at all. This could be bundled together as a book, a paper drawing, a letter but also bundled together as a digital format like an eBook, a CAD file, an email and even transmission bytes of a network / internet provider can be considered as data.
Data becomes significant once provided in the context of each other or in the context of other data. At that time, we start calling it information. For that reason, a book or a drawing provides information as the data has been structured in such a manner to become meaningful. The data sent through the network cable only becomes information when it is filtered and stripped from the irrelevant parts.
Information is used to make decisions based on knowledge. Knowledge is the interpretation of information, which combined in a particular way, helps us to make decisions. And the more decisions we make and the more information we have about the results of these decisions, either by us or other, it will increase our knowledge.
Data and big data
Now we have some feeling about data, information and knowledge. For academics, there is room to discuss and enhance the definition. I will leave it by this simple definition.
Big data is the term for all digital data that is too large to handle in a single data management system, but available and searchable through various technologies. Data can come from any source around the world as through the internet an infrastructure exists to filter and search for particular data.
By analyzing and connecting the data coming from these various sources, you can generate information (placing the data in context) and build knowledge. As it is an IT-driven activity, this can be done in the background and give almost actual data to any person. This is a big difference with information handling in the old way, where people have to collect and connect manual the data.
The power of big data applies to many business areas. If you know how your customers are thinking and associating their needs to your products, you can make them better and more targeted to your potential market. Or, if you know how your products are behaving in the field during operation (Internet of Things) you can provide additional services, instant feedback and be more proactive. Plus the field data once analyzed provide actual knowledge helping you to make better products or offer more accurate services.
Wasn’t there big data before?
Yes, before the big data era there was also a lot of information available. This information could be stored in “analogue” formats ( microfiche, paper, clay tablets, papyrus) or in digital formats, better known as files or collections of files (doc, pdf, CAD-files, ZIP….).
Note the difference. Here I am speaking about information as the data is contained in these formats.
You have to open or be in front of information container first, before seeing the data. In the digital world, this is often called document management, content management. The challenge of these information containers is that you need to change the whole container version once you modify one single piece of data inside it. And each information container holds duplicated information from a data element. Therefore, it is hard to manage a “single version of the truth” approach.
And here comes the data-oriented approach
The future is about storing all these pieces of data inside connected data environments, instead of storing a lot of data inside a (versioned) information container (a file / a document).
Managing these data elements in the context of each other allow people to build information from any viewpoint – project oriented, product oriented, manufacturing oriented, service oriented, etc.
The data remains unique, therefore supporting much closer the single version of the truth approach. Personally I consider the single version of the truth as a utopia, however reducing the amount of duplicated data by having a data-oriented approach will bring a lot more efficiency.
In my next post, I will describe an example of a data-oriented approach and how it impacts business, both from the efficiency point of view and from the business transformation point of view. As the data-oriented approach can have immense benefits . However, they do not come easy. You will have to work different.
Some more details
An important point to discuss is that this data-oriented approach requires a dictionary, describing the primary data elements used in a certain industry. The example below demonstrates a high-level scheme for a plant engineering environment.
Data standards exist in almost any industry or they are emerging and crucial for the longevity and usage of the data. I will touch it briefly in one of the upcoming posts, however, for those interested in this topic in relation to PLM, I recommend attending the upcoming PDT Europe. If you look at the agenda there is a place to learn and discuss a lot about the future of PLM.
I hope to see you there.
2 comments
Comments feed for this article
August 19, 2014 at 9:11 am
youhey
This post is really a nice and succinct problem definition of files- and data-oriented approaches.
There exists already various solution that replace MS-Office for different activities during the product development (from Idea management and Requirement management to sales and service). They are also partly integrated in PLM systems. However, everything will be more complicated as we move from MS-Office based activities towards ECAD, MCAD, calculation, programming, …
I think, a hybrid solution with semi-automatic data extraction can be practicable and acceptable in short- and midterm. Also a model based product development will be probably necessary in order to take full advantage of the hybrid solution.
PS: A nice comparison of Information and Knowledge:
“Newton’s second law of motion says that, the sum of all forces on any object is equal to the mass of that object multiplied by its acceleration (F=ma). In this example m (mass) and a (acceleration) are two knowledge elements (information) and their multiplication is the knowledge process to create another knowledge element F (force).”
Based on this, information is just a knowledge element and new information is result of a knowledge process in a predefined context.
Greetings,
Thanks Youhey – not sure if I understood the F=m.a analogy, but I might be blurred by my own physics study. Best regards Jos
LikeLike
August 27, 2014 at 12:27 am
Henk Jan Pels
Hello Jos,
Very interesting subject. There has been a lot of discussion about the definition of data, information and knowledge and there are quite different interpretations, but years of experience in IT lead me to one that does not claim to be more true (even rather narrow), but only to be more practical in the Information Systems context:
• Data is a (set of) statement(s) about some fact in a relevant world,
• Information is data that is relevant for some decision to take
• Knowledge is the ability to perform a task and is related to information in the formula:
Knowledge = Skill * Experience * Information * Attitude (Weggeman)
Opposed to your definition data must have a meaning by being coded in some language. Patterns on paper that have no meaning are not data but noise. Also the word data is linked to the “givens” in a mathematical problem: statements about the environment of the problem. If I type a question in Google the response is data: a set of statements. Hopefully I find a statement that holds (part of) the answer to my question. That statement is then information for me in that specific context. The purpose of a business information system (like PLM) is to manage data (e.g. statements about products) in order to be able to make the proper selection to provide information to persons that have to take some business decisions.
You propose that knowledge is the interpretation of information. My problem is that interpretation is already in data: if you do not understand the language in which the data is coded, it is just noise for you and cannot be interpreted as a statement about a fact you might be interested in. Where business is about designing, making, selling and supporting products and services and is organized in a processes as networks of tasks, the quality of performing those tasks is very important for the success of the business.
Knowledge is what makes or brakes this quality and PLM is about bringing the right knowledge to the right person at the right time to perform a specific task. Information is one part of it: skill (to be learned from book, courses, examples or trial), experience (progress in the learning curve) and Attitude (willingness to perform) are just as important. If we take it as the task of PLM to organize vthe most effective flow of knowledge, then these are practical definitions to make PLM understandable.
Henk Jan, thanks for your clarification, which extends the definition of data, information and knowledge. For sure when we meet again we can have a proper discussion as you might realize I try to simplify concepts although I feel it is still complex to explain even using simplification.
LikeLike