Archive for the ‘Blogs’ Category

Doing carbon transparency right – here’s how

Publicerad: 11 januari, 2023 av Alexandra Jerrebro |

In 2019, Emma walked towards the meeting room that she had booked for the entire day. She only had one point on the agenda:

  • Environmental Goal 2030 – Enabling our customers to be carbon neutral

“There is a high risk that we spend the entire day just discussing and we won’t get any closer to any type of action”, she thought as she entered the room. A few of her colleagues stood by the pastries and sipped coffee discussing a particular customer challenge they were working on.

“This is the third major customer this quarter who’s been requesting more granularity in traceability and carbon footprint. We’re already ahead of our competitors in carbon emission data, how can they still ask for more?”

Emma joined the conversation, “What if we did something bold, what if we had full transparency from cradle-to-cradle. Feedstock, transportation, processing, distribution, recycling – the works?”.

“It’s impossible to be that granular”, another colleague pitched in immediately.

“In this room we have some of our best experts in purchasing, supply chain and processing. Are you saying that we have a challenge in front of us that we’re not able to solve as a group”, Emma challenged him.

“Let’s begin the meeting”, Emma then announced and quickly started the meeting with this slide.

What is your biggest hairiest problem?

When I start working with my clients, I ask them:

“What’s your biggest hairiest problem?”

The answer might be something like this:

“We want to create full carbon transparency end-to-end for our products. But it’s impossible.”

My follow up question is:

“What data, information and knowledge do you need to solve this problem?”

“I need to know the carbon footprint for each type of feedstock, which is something our suppliers don’t have. I need to have traceability of feedstock throughout our enrichment processes, something which we can’t do. I need to know what the carbon footprint for transportation across 5 continents with 100s of transportation companies. I need to know the energy usage for the processing of groups of feedstock in combination with the local site energy mix. We have none of this today. I don’t even know where to start.”

Narrowing it down

Let’s break down the situation in its individual pieces.

For this specific case, we need to create a data model which supports the following:

  • Feedstock Types
  • Relation between Feedstock Types and Product
  • Product
  • Transportation Methods
  • Relation between Transportation Methods and Feedstock Types and/or Product
  • Organization
  • Relation between Transportation and Organization
  • Production process
  • Relation between Production Process and Feedstock and Product
  • Relation between Organization and Organization through a Business Party relation
  • Business Party
  • Geography
  • Relation between Production process and Geography
Figure 1: All of these elements need to be able to hold CO2 emission values. Feedstock and Production process can be reused to model the Energy Mix for an individual site.

(Click on image to enlarge)

Get to work

Once you’ve narrowed down the issue in its individual parts, it’s time to get to work.

A typical Product Model looks like this. It says that (reading from bottom to top):

  • A Product Individual (has its own serial number, can be bought by a customer) can consist of another Product Individual.
  • A Product Individual is related to a Product Article. A Product Individual is a produced based on the Product Article’s specification. The Product Article’s MBOM creates the first Bill of Material for the individual but then the individual follows its own lifecycle, i.e. it may get new parts with new capabilities.
  • A Product Article is the specification of the Product. It holds all the attribute values for the product. Product Article is what is specified by engineering (EBOM – As Specified) and what is produced by the factory (MBOM – As Built).
  • A Product Article (shares an article number with other Product Articles but is not an individual product which a customer can buy).
  • A Product Article is related to a Product Concept and to a combination of Organization and Role.
  • A Product Concept is a Generic Variation of a Product.
  • A Product Concept is a generic description of the product which is stable over time even if a new version of a Product Article is released under a new article number.
  • A Product Concept is related to a combination of Organization and Role.
  • Generic Product is the master classification structure which holds the definition of which Attributes are defined for each Product. 
Base Model: Product
Figure 2: Base Model: Product

(Click on image to enlarge)

The key in solving complex data issues is to build out the data models step by step whilst ensuring connectivity to other data models. To capture Carbon Emissions the product model was extended on the article level to capture the as-built BOM.

Click on images to enlarge

The model is continuously evolved, expanded and tested. Ensuring that use cases and data match are supported by the model.

An Operational Reference Model

Carbon Transparency is no different. Once you have the data models defined, it’s necessary to visualize, validate and improve the data models with actual data. This is the core of what we expect from a Digital Twin Platform. This increases the demands on a data model as it needs to be live, executable and populated with real data. I call it the Operational Reference Model. If this is achieved, it will support the following functions:

  • Centralized and distributed governance
    • Support both centralized and distributed governance of meta and master data.
  • Connectivity
    • Be agile in its relation to existing solutions, to supports both existing meta and master data management solutions as well as covering gaps in those systems.
  • Hold and maintain Taxonomy and ontology for increased data quality
    • Taxonomy and Ontology are key elements for creating the necessary context for any AI and ML initiative. A platform designed to build, maintain and distribute taxonomies.
  • Active meta data management
    • To manage and maintain simple-to-complex meta data which is used in PLM, ERP, CRM, PIM, PDM and MDM as a common language for data exchange.

New capabilities

Going back to Emma, what happened after her full day working meeting?

She had an ace up her sleeve, as she already had an Operational Reference Model in place for Product which gave her a head start. Before the end of the year, she was able to deliver new capabilities to the organization:

  • Supporting the business with environmental information to be able to give correct information to customers.
  • Supporting the consumer demands connected to environmental information.
  • Be able to simulate (future demands on) product configuration in an environmentally friendly way.

 The 2030 vision that seemed impossible at first, was now perceived as achievable. One step at a time.  The requirements were in place, the roadmap to achieve it was set. Suppliers were informed of data requirement ramp-up the coming years. The key was envisioning the future, taking a future back mindset, and going backwards to define what is necessary today.

To do carbon transparency right is possible. What happens if you don’t do it – for your own organization or your customers?

/Daniel Lundin, Head of Product & Services

Daniel Lundin Ortelius digital twin

If you had to do a Football World Championship 2.0 model

Publicerad: 14 december, 2022 av Alexandra Jerrebro |

An information modeler’s challenge – how would you have done it?

How is a Football World Championship really structured and what are the components required to make up the whole? Football associations, national teams, players, referees, coaches, equipment, assets such as stadiums and not least the games themselves. So, what to do with a challenge like the Football World Championship? A question the international governing body (with whom Ortelius has no commercial association) have asked and answered with, at best, mixed results in recent years. It is also a question that we at Ortelius found ourselves standing in front of a few weeks ago, at least from an information modelling point of view.

Like most workplaces, we like to have a football prediction competition during major championships. However, being information modellers, we could not resist the temptation of stretching the parameters of our competition. Could we model a football competition and break it down into it parts? And even better, could we build a model that would be sustainable enough to extend to other competitions and live into the future? After all, this is something we do for every major championship.

But where to begin? An issue which confronts us not just in the case of this model, but in any model for which we start with a blank canvas. The problem with the sky being the limit is that the sky is intimidatingly big, and a limit, initially at least, would often be quite welcome. We address this by defining a use case against which we can develop our model. What is key to remember here is that the model is not only built to solve this use case, but rather the use case gives us a frame against which to set the scope of our model. We need to define all the entities needed to solve our problem and provide immediate value, but we want to do this in such a way that those same entities can be used when additional use cases come.

Our initial use case was of course our prediction competition. In our competition at work, all competitors would predict the results of all games and would receive points for guessing the correct result with bonus points for guessing exact scores and correct goal difference. To achieve this, we would need to model the teams to compete in the games, the games themselves, their results, and the groups and knock-out formats in which the games would occur.

Figure 1: Draft of information model

(Click on image to enlarge)

Our first step is always defining the taxonomies in our model. Creating detailed taxonomies allows us to create a surrounding ontology with connections at the right level. We can use our definition of team as an initial example. Brazil is an iconic and enduring presence in world football but current team of Neymar and Richarlison is not the same entity as the team of Rivaldo and Ronaldo. This is also the Men’s Football World Championship; Brazil has a Women’s team which would also need to be supported by this model in the future. Remember, we are building a model not only to solve our current use case but future use cases too.

In addition, Brazil is a national team with a very distinct ontology from a club team. It competes in international competitions against other national teams rather than in a domestic league against other clubs. From this, our taxonomy emerges distinguishing first between football teams and, for example, rugby teams and then branching out with club and national teams. We can then create our variants of national team such as Senior Men and Senior Women. Finally, we can create our individual teams based on these variants such as the current Brazil squad containing Neymar and Richarlison. The current Brazil squad can be connected to the Qatar Football World Championship while the concept of the Brazil national team can be connected to the Brazilian Football Association. This may seem self-evident in this case but a poorly defined taxonomies can be found at the root of a substantial number of data problems.

The next challenge was to define a game of football. Our starting point was to try to define it as we would any activity such as a task or manufacturing process. A game has a defined length in time, resources (players, officials, stewards, etc.), equipment (goals, footballs, cameras), an output (a result), and location (stadium). We developed our game taxonomy to differentiate between football games and other sporting games as this gives us a frame to differentiate in terms of attributes such as game length, but also a clear structure with which to connect our football games to their wider ontology. For example, group games are 90 mins and included in a group while a quarterfinal will have the potential for extra-time and penalties and will be followed by a semi-final. We can then begin our ontology work by connecting the teams we defined as a resource which enable the game much like we would connect a machine to a manufacturing process. There is a certain satisfaction in using modelling concepts that you would usually use to model a production line in a factory to solve the problem of France vs Tunisia.

Figure 2: Information model in application

(Click on image to enlarge)

We then continued our taxonomy work defining groups and how they relate to teams and the competition as a whole. A proposal has been floated to reduce the groups to three teams in 2026 and expanding the competition to 48 teams so the nature of this taxonomy and ontology could change significantly in the future. And finally, we have our predictions. The games have an actual result of course but it is important to capture our competitors’ predictions. Again, this is much like any activity model with a forecasted and actual result.

We are now coming to the end of our competition and by populating and testing our model with data we have learned a lot about its strengths and weaknesses. Our taxonomies could do with some refining to allow better inheritance for both teams and games but that is the nature of taxonomies, they are never truly complete and need to have the flexibility to be evolved based on testing and new information.

The real challenge will come with the next tournament with which we want to test the model and we will see if it is resilient enough to embrace that one as well, but for now we should get back to doing some actual work.

(Click on images to enlarge)

Figure 3: Information model in application

Figure 5: Example of Group Tables

Figure 6: Dynamic digital model made in inorigo

/Ferdia Kehoe, Senior Information Modeler

A new breed of business consultants of 2023 will know data and information modeling

Publicerad: 7 december, 2022 av Alexandra Jerrebro |

Let’s start with a couple of questions to set the scene.

  1. Do you believe data and information are important parts of developing your business?
  2. Do you believe that decisions you make today will have an impact on the success of your business tomorrow?
  3. What are the first steps when you start developing a new product?

As a business consultant you are often faced with a wide array of tasks and topics. Many (most) include data in one form or another. Increased specialization and the drive to aggregate expertise, assign accountability and providing a sense of identity have led to increasingly siloed organizations, as discussed in the November issues 2021 of Harvard Business Review, HBR1. Organizations are adopting different methodologies (e.g., SAFe Agile, Hub and Spoke, Teal organization) to manage this and increase knowledge sharing. Increased knowledge sharing across silos enables employees to sell more and learn more.2

This brings us back to the question of the new breed of business consultants and their knowledge of data and information modeling. Can they help companies in sharing information across silos, enable growth and boost learning? And how can a business consultant provide a framework for knowledge sharing in a way which makes sense to the recipient?

A new breed of business consultants works with these questions and think differently. Consider following statements in relation to the questions previously asked.

  1. When you develop your business and its capabilities you should ensure that data and information are developed and structured together with the business.
  2. What data model you create today affects what generates the data and information that you need to make decisions on in the future.
  3. When you start developing a new product or construct a new building you often start with an architect/designer because they are able to design and visualize the result. This in turn ensures that everyone is aligned on the result before you start. The same goes for data and information when you work with strategic decisions changes in your organization.

If you want to ensure that all new capabilities will generate data which forms the foundation for your ability to gain insights, share knowledge and increase sales in the future, you will need to design your data model alongside your operating model. Or rather, as a part of your operating model.

To be clear, a data model in Excel only takes you to base camp. It’s a great starting point, but you will require additional skills and preparation to reach the summit.

One way of considering the role of a data model in the everyday life of a business consultant is talking about semantics. Revenue and Product are two terms often used when you work with a business consultant. But are we entirely clear on what aspects of a product the person we talk to use?

A bigger question here is: Do we talk about a Product from an Engineering, Financial or Sales perspective? For products, a Taxonomy (a logical, hierarchical structure) and an Ontology (how a Product and an Organization are connected) provides a way of establishing a language where it’s clear what context we are operating in.  

Figure 1: Different levels of Product and functions in an organization

(Click on image to enlarge)

Taxonomy and Ontology are the two primary tools in the information modeling workbench that propel a business consultant close to the summit – and consequently close to summit for the customer. They allow people and functions to understand the relation between themselves. Once established, it allows data to be structured and used cross functional. A Taxonomy and Ontology works independently of systems, whilst still enhancing the capabilities of individual systems. It supports processes and provides clarity in what data and information is enriched and maintained in each step of the process. A taxonomy gives a solid framework for business, process and data to work jointly.

Well designed and put in operation, it will ensure ultimate adaptability for an organization who want to evolve over time!

Figure 2: A taxonomy allows for common understanding of the different aspects of a Product

(Click on image to enlarge)

An example of this is a large manufacturing company which we at Ortelius work with. In one case, the customer was about to launch their first eBusiness solution, but they had two major concerns that caused inefficiency. 1) There was no single source of truth for their products, one and the same product could have up to seven different names depending on who you spoke to in the organization, 2) They wanted to be customer centric and thus not create an eBusiness solution based on their internal article number database (2.5 million sales items). Together with the customer, a Product Taxonomy was developed and designed outside in (“how does the world and our customers see the products we produce”-perspective) and created a single source of truth for a commercial product offering. The commercial Product Taxonomy then had relations to one or several articles rendering the ability to sell and market the same article under different names/brands in all customer facing channels without creating data redundancy across the entire data landscape. Once this was in place it was easy to start building knowledge around the products. Marketing, Sales, Patent/Trademark, Operations, Supply, Service and Product Management all had different needs on product information, but also product information from other departments. The solution enabled Service to automatically obtain information from Sales about which industries and markets the product was sold in – this improved insights and customer service efficiency, Product Management obtained information from Patent department about which products had IP protection – this minimized business risk, Marketing obtained information from Product Management about which translations were approved – this improved sales channel efficiency and time to market.

Traditional Business Consulting consists of gathering data and information, collating it and presenting it and/or implementing it to improve efficiency or increase sales for a customer. The work on structuring data and information is being done anyways, but customers are beginning to ask for more than a PowerPoint or Excel delivery. Customers are asking for a sustainable, governable way of solidifying the work already done. If the data model is created, why not take it from base camp to the summit and ensure the customers receive the value they need. This is the role of Business Consultants 2023 and onwards.

/Daniel Lundin, Head of Product & Services

Daniel Lundin Ortelius digital twin