The difference between files and data-oriented – a tutorial part 2

In my previous post, I talked about the unstoppable trend towards digital information and knowledge based on data becoming the new business paradigm.

Building knowledge based on information extracted from data, instead of working with documents and people, who need to manipulate these documents.

Moreover, the reasons to move towards a digital data-oriented approach are the immense business benefits it can bring to an organization. Having online visibility on information in context of other information from different stakeholders allows companies to be more proactive.

A proactive company will react faster to the market or customer. This will reduce waste and resources (materials / people) and therefore in the end be more competitive. This is all described in my first post, with relevant links to various global references.

In this post, I want to describe in an example what the differences are between a document-oriented and a data-oriented approach and how it affects people and business. This might give you an impression of the expected business benefits.

The ultimate goal behind a data-oriented approach is to have a single version of the truth for a product, project or plant. This can be realized by treating information as data elements in various connected database, where on demand reports or dashboards can be created based on actual information, instead of documents generated by duplicating data in new systems and locations. Digital data will provide paperless processes accessible almost anywhere around the world.

As an example, I will explain the difference between document-centric and data-centric when dealing with specifications.

The specification

Everyone knows the challenge with specifications. Most of the time in printed documents describing how a product or service should work from the client point of view. There are two principles behind specifications:

• Complexity. The more complex product or service is, the bigger chance that specifications are not complete or hundred percent understood, leading to an iterative change process. The challenge here is to manage the change and the consistency of the full specifications.
• Industry and margins. In a repetitive business, for example, the automotive or other mass consumer products, products can be quiet complex and once sold hard to maintain and repair. In a competitive business, an error in the field can consume a lot of the expected profit. In the construction industry, where most of the time single projects are executed by a chain of disciplines, the industry (still) accepts the costs overrun and the high costs of fixing issues in the field, instead of being clearer upfront during the design and planning.

Let’s stay with the example in the middle of complexity and industry volume. In color the various stages of the process.

The document based specification – 1

When the document based specification arrives, the company has to get an understanding of the content. The project manager has a first read through the document (100+ pages) and decides to send the document (it is a pdf) to sales, engineering, legal and planning. Engineering decides to distribute the document internally to mechanical, electrical and quality (for compliance).

The project manager stresses everyone on a weekly base to deliver the responses and tries to understand if the answers will come in time. There are some meetings needed with the stakeholders as the whole understanding needs to be consistent. Based on several iterations a response is compiled.

The data-oriented approach – 1

When the document based specification arrives, the project leader first stores the document as a reference in the PLM system and extracts all the customer requirements as data elements in the system. While extracting the requirements, the projects manager groups them into digital folders (functional / non-functional, contractual, regulations, etc.) and assigns them to the relevant stakeholders, who get notified by the system. Each of the persons assigned, again the engineering manager has distributed the discipline specific requirements internally.

The project manager watches the progress of the requirements analyses which are around a virtual model. There is still a need for meetings with the stakeholders to agree on the solution approach. Everything is stored and visible online in the system. This visibility has helped some of the stakeholders to be better-aligned upfront. In the end, the response is generated and converted to the customer’s format.

Not much benefit for step 1

If you compare the two approaches, there is mainly one person happy: the project manager. Instead of spending time to collect the status of all information, direct visibility on the response helps him/her to prioritize of focus where attention is needed, instead of discovering it on a weekly base.

There is some small benefit from the virtual model as other stakeholders can have a better understanding of the actual progress.

However, for the rest, all stakeholders are complaining. It is difficult to work. (Fl)Excel was much easier. Moreover, thinking about a virtual model takes time as we are not used working in this way. Typically something for aerospace you might think.

And now the benefits come – step 2

The customer has placed the order, and the project has started. The design has started, and people start to discover discrepancies or ambiguous demands that need to be negotiated with the customer. Is it part of the project and if not, should it become part of the project and at which costs (for whom)

The document oriented approach – 2

Several engineers are now discussing with the counterparts at the customer the detailed interpretation of the requirements, either through face-to-face meetings or emails. Changes are collected and sent to the project manager, who tries to understand what has changed and how to merge it in an on-going specification document. To avoid many revisions, he/she tries to update the document on a bi-weekly base, send it to the internal stakeholders for review and with their feedback generates a specification document for the customer that supposed to cover the latest agreements.

Unfortunate not all changes have reached the document as some of the stakeholders were busy and forgot to include some of the changes agreed with the customer as they were in a lost email. Also, a previous change of a requirement was overwritten as an update from quality used the old data. Finally, some design solutions were changed, which raised the costs. And not sure if the product with all its changes will be compliant after delivery. However, luckily nobody noticed so far, not even the customer

The data-oriented approach – 2

Thanks to the virtual model and the relations between all the requirements, any change in a requirement gets notified in the system. When a requirement is further clarified, it is updated in the system. When a requirement needs to be changed, it is clear what the impact of this change is. A change workflow assures that decisions are made visible and approved. Potentially changes that lead to more work were quoted to the customer for acceptance. Luckily the compliancy engineer noted that the change of materials used would lead to a compliancy issue. On a bi-weekly base, the project manager generates an agreed specification for the customer based on the data in the system.

Benefits are growing.

The project manager remains the happiest person and is even happier as less discussion is needed about who changed what and why. Alternatively, discussions about changes that should exist and cannot be found. The time saved by the project manager could be used to collaborate even better with the teams (without annoying them) or perhaps a second project to manage in parallel.

Other stakeholders start to enjoy the data-oriented approach too. Less ambiguity on their side too, fewer iterations because changes were not apparent. As all information is related to the virtual model online, the actual status is clear when making a decision. Less fixing afterwards and luckily still project meeting between the stakeholders to synchronize. The PLM system does not eliminate communication; it provides a reliable baseline of the truth. No need (and option) to look in your archives.

At this stage, benefits start to become clear. Fewer iterations and better decisions will have an impact on the costs and project stress. Still a complaint from the engineers might be that they need to do too much upfront thinking although some years later they might discover that this will be their main job. Fixing issues from the past have diminished.

And then the ultimate benefits

Now the project has reached the physical state. It is manufactured or under commissioning.

The document oriented approach – 3

In the document oriented approach, many issues might pop-up because they have not been considered in the early phase, or they got lost during document exchanges. Does the product work as specified? Is the building certified as specified?

The customer is king and for manufacturing companies this might lead to product recalls or launch delays. In the construction world, people in the field, will fix the issues by using skilled resources and creating a waste of materials and/or resources.

Data handover to the owner is a nightmare for a the project-centric delivery. Several people have been searching for documents, specifications and emails to build and compile the required documents for handover.

The data-oriented approach – 3

In the data-centric approach, the behavior of the physical product works as expected as most of the issues have been solved in the virtual model. When testing the product it works as specified as the specifying requirements have always been linked to the product. Moreover, they have been agreed and approved by the relevant stakeholders. Where relevant, the customer has paid for the extra work specified.

The handover process was not so stressful as before with the document-oriented approach. As the required information was known and specified upfront related to the requirements, the maturity process of the virtual model assured this data exists in the system. Now the as-built information matches the as-specified information. What a relief.

Conclusion

It is clear that the significant benefits can be found in step 3. I wrote the comparison in an extreme manner, knowing that reality lies in the middle. Excellent people can comprehend and fix more upfront because of their experience. Building the ultimate virtual model is not yet an easy achievement either.

The savings in materials and required resources are significant in a data-oriented approach. The time savings and the quality enhancements might change your company into a market-leader. The cost savings achieved through a pro-active approach will make your margin growing (unless competition does the same) and enable you to innovate.

One final remark on business change

If business change could be achieved by selecting the right tool or system without a business change, you will never get a competitive advantage. As your competitor can buy these too.

However, if you change to a data-centric approach, it will be a though change process and therefore once implemented you will leave competitors behind that keep on hanging on the past.