From Dirt to Data

There is economic value in connecting people to information; improve the connections and you increase the value of the information.

Consider the analogy in which your manufacturing or construction project is a ditch, and data inside your files is water that will flow through that ditch. What sort of tool are you using to move the dirt and prepare for the water, shovels or backhoes? I would submit that Windows Explorer is the equivalent of a shovel, and product data management (PDM) software is the backhoe. Product Data Management (PDM) moves a lot of data fast. And it does much more than that. PDM also improves access, reuse, sharing, record keeping, and security that affect data files.

After digging the ditch, you have a nice place to store data. But that place to store data is not a static repository. It is a moving river, where data flows freely. Data is created by CAD users, deposited into the river, and then flows downstream to manufacturing shop floors, documentation departments, and customers, and other consumers of the data.

When your data management metaphor is a river, you can gain valuable insights into managing the data in your organization. These four V’s demonstrate things about the rivers of data that can help you understand and solve some of the data problems in your organization.


In 2011 the average US manufacturing company had 300 times more data stored than it did in 2005. Globally, computer users are generating 2.5 quintillion bytes of data annually; 90% of the world’s data has been created in the last two years. No matter the size of your company, you have a lot of data and it is growing at a rapid rate. Getting control over it is critical for your survival.


There are numerous file types associated with each major CAD platform, and it is the rare organization which has the luxury of dealing with only one CAD format. You need a system that can manage the complexities of SOLIDWORKS or Inventor as well as MicroStation or AutoCAD. Each one has peculiarities that require a unique approach.


At the corporate level, consulting firm McKinsey claims poor data quality costs the US economy around $3.1 trillion per year. In product design, data quality is critical. Sending the wrong model or drawing to a subcontractor is more than a waste of time; it can cause significant cost overruns if not discovered quickly. Multi-CAD projects require management of interoperability, with opportunity for errors to creep in if not addressed proactively.


When measuring the time it takes to manage good data, you have to take into account many operations, all of which take time. You have to search for existing models, or parts, or revisions. You have to ensure you are working on the correct version. You have to assign tasks to other workers, and receive assignments, as well. You have to be sure documents are accessible to other users, including publishing finished documents so they can be easily found. You have to make metadata about the document accessible. All these operations go faster when using product data management (PDM) software.

When you look at the mounds of data piling up around you in your organization, you should seriously be looking for the best tools to organize this data and provide you with the most useful data stores possible. Synergis Adept product data management is an ideal tool to help you leverage your data to make the most of your business.

Randall S. Newton is the principal analyst and managing director at Consilia Vektor, a consulting firm serving the engineering software industry. He has been directly involved in engineering software in a number of roles since 1985. 

3 thoughts to “From Dirt to Data”

  1. I’ve had the opportunity to manage an Engineering Data Cleaning project. The company (a global manufacturer) had the feeling that the low quality data they had cost them a lot (and in fact we identified over 14,000 potential item duplicates and 12,000-15,000 part numbers that could be removed through simple standardization). At an estimated cost of $450/item/year for moving items, that’s a lot of money the company was wasting.
    The project was done mostly through manual ETL, but after it ended I managed to automate the extraction (for future opportunities – over 50% of the manual work can now be eliminated).

    The issue is that usually the cost of bad data is not visible (being buried into the operational costs) and the bean counters’ radar does not pick it up. Any project that attempts to deal with the issue needs to quantify the hidden losses and show the clear numbers in order to justify the budget.
    From what I also learned is that companies try small data cleaning projects – here and there, in one department or another, without a holistic approach. Almost all of them fail, making the “data cleaning” concept something similar to “failure”.
    So the $3.1 trillion price tag seems a good estimate…

    1. Thanks for your comment.

      What’s interesting to me is that in over 20 years of selling engineering data management solutions, there are just a handful of companies that are able to calculate the amount of waste due to bad data. The examples that come to mind are Scott Paper (now Kimberly Clark) who spent $250,000 creating a prototype for a paper machine only to discover it was created from the wrong version; and Taggart Global (, that spent $760K per year on manual drawing registers and transmittals.

      The cost of waste, whether due to bad data, the variety or veracity of data, is astounding. And the fact that so few companies see that “getting their data in order” as mission critical is also surprising, given that engineering and product data management solutions have been around for a couple decades.

  2. Florin, I agree in principle with your comments. But the specifics of the situation are very different depending on the size of the company and the nature of the work. Project-based engineering (construction, infrastructure, etc.) has different issues from manufacturing.

Leave a Reply

Your email address will not be published. Required fields are marked *