Poor data integrity in a set of drawings happens when there is inconsistency between the descriptive information shown in a drawing’s title-block (i.e. drawing number, revision, revision date, title, status, etc) and the information (metadata) used to tag that drawing in a document management system or project extranet. In other words, what you see may not be what you get.
Poor data integrity causes many problems. Here are a 3:
- Most frustratingly: people can’t find what they’re looking for! It’s the garbage-in-garbage-out principal.
- Inaccurate reporting. If a set of search results doesn’t accurately reflect what it claims to reflect, trust in the system breaks down.
- Wasted time and money. Document control teams can waste weeks simply checking the accuracy of a submission. Error rates of 3-5% are standard. On a submission of 5000 drawings, that’s 150-250 mistakes. Then there’s time taken to correct those mistakes. On critical submissions this can delay the whole project.
So why does it happen? Here are the 3 main reasons:
- Manual processes dominate the way drawing information is managed. Drawing lists are developed by hand, often cutting and pasting from previous jobs. Document controllers and CAD operators are forced to manually enter data into title-blocks as automation functionality is limited.
- Drawing information is fragmented. Information is being authored by multiple organisations and replicated across multiple mediums. Excel drawing lists, CAD files, file names and project extranets all contain descriptive information about drawings. Keeping one set of data consistent across all these mediums is difficult.
- There’s no single source of truth for drawing data. Extranets are good at managing completed documents, but offer little help to architects, engineers and contractors in the authoring process prior to their drawings becoming drawings.
The industry desperately needs a new solution. A solution that bridges the CAD/BIM environments in which designs are authored, to the project extranet / DMS environments, where drawings are uploaded, indexed and distributed. One key to this solution will be the centralization of drawing data, such that there is a single source of truth. The other key will be that systems talk to each other such that the correct data can reliably move from one environment to the other without the need for error-prone humans to do it by hand.
At DAD we’re passionate about this stuff. This need is a driving motivation behind the work we’re doing in the development of the Database Aided Design platform.