I’ve spent most of my working life managing information on projects. One frustration I’ve witnessed on an almost-daily basis is the prevalence of poor data integrity in construction information, and the crippling impact that has on projects.
Example. The Project Manager searches the project extranet for all drawings tagged with the status Issued for Construction. They transmit them to the Contractor. The Contractor prints the drawings and uses them to order materials and prepare shop drawings. Weeks later it is revealed that the Architect incorrectly tagged several of those drawings in the extranet as Issued for Construction. Printed in the title block, the status is Issued for Approval. The contractor has now spent money and wants to be paid for materials that were actually never approved. And so the finger pointing begins. All because of poor data integrity!
In another recent example I witnessed, an Owner of a $600m project was unable to easily insert an accurate drawing register of all design drawings issued during the tender phase into the main contractor’s contract. The problem was that the data in the drawing register (containing over 5000 design drawings) extracted as an excel report from the project extranet didn’t accurately match (i.e. drawing number, revision, status) the drawing registers of the various design consultants, and in many cases the data in consultant’s own register didn’t accurately reflect the data inside the drawing title-blocks. Days of manual checking revealed hundreds of mistakes, which delayed the start of the project.
Construction and engineering projects generate oceans of information. The growth of web-based document management systems (DMS) and extranets over the past decade has gone a long way to provide much needed structure and searchability of project documents. Yet the people I talk to still complain about not being able to find what they’re looking for.
Sometimes a lack of training (or more correctly, a lack of willingness to be trained) in the functionality of the project extranet can be blamed for why people struggle to find what they’re looking for. However these days, anyone with half a brain can quickly learn how to navigate through what are generally well-designed systems. The real problem is not in the searching; it’s in the data integrity.
No matter how good the extranet system is, there’s no escaping the “rubbish-in-rubbish-out” principal. Ensuring documents are labeled with accurate and descriptive tags is the key to achieving Data integrity. Here the 3 steps on how to do it.
First and foremost, project teams must document well-defined protocols around document numbering, categorization and uploading procedures. The earlier in the project this happens, the better. Day one of the project is ideal so make sure it’s on the agenda. The protocols must not be overly complex because “over-categorization” does more harm than good. Start simple, and then develop the protocols as the project requirements evolve.
Secondly, explain and enforce the protocols. High-level buy-in and support form the Owner and the Owner’s Representative (Project Manager) is key. Relaxed enforcement sends a message to the project team that the protocols are mere “friendly suggestions” when in fact they should be understood as law. If you take the time early in the project to properly (and patiently) explain protocols (and indeed incorporate any feedback) the project team both respect and appreciate the upfront clarity of what is expected.
The third and, to me, most interesting challenge in achieving real data integrity, is that of keeping one set of data consistent across multiple mediums. On today’s projects we encounter this challenge when architects, engineers and contractors try to ensure that the information in their drawing lists are accurately reflected in the title-blocks of their actual drawing as well as the metadata in the project extranet. Too frequently however, we see information updated in one location but not all locations, resulting in inconsistency and unreliability like in our example above.
To rise to this challenge, teams must adopt new technology, like Database Aided Design, which establishes a “single source of truth” for drawing data and eliminates the need for manual data manipulation across mediums.
In my next post, I will dive deeper into this last point as it deserves a lot more attention.