Comments on: Data Quality and PLM

By: olegshilovitsky

olegshilovitsky — Wed, 16 Dec 2009 18:04:08 +0000

Swati, The biggest problem I see is that nobody in the company is taking responsibility for data quality. This topic is falling between chairs of engineering and IT… I understand, you can try to clean data when moving to a different environment, but this is a very expensive task. Most of the companies are struggling just to import what they have outside. Best, Oleg

By: Swati Ranganathan

Swati Ranganathan — Wed, 16 Dec 2009 17:43:38 +0000

Oleg
this is an interesting set of questions you pose…in most cases, data quality within ERP/CAD systems is poor. So as a result, PLM inherits a mess, unless a lot of time is spent on cleaning up data (which is rarely done).

As part of my implementations and data migrations, I have tried to clean up data as much as possible so that structured information within PLM can then start to contribute to efficiency gains.

Since most of the implementations are initiated and managed by the engineering department, data quality is the last thing on their mind. Design quality and time to market is…Very often this falls on the IT / business analyst to clean up/manage as a hobby and not their job function.

By: olegshilovitsky

olegshilovitsky — Tue, 08 Dec 2009 21:57:12 +0000

Cam, thanks for comment! By definition, data managed by one system is in pretty good shape. Problems come when you have multiple systems and that what happens on in organizations. If part information in ERP is different from PLM, you have a problem. What PLM can potentially do is to federate/manipulate data from multiple systems. If this is what implemented in the organization, you have better data visibility, and it will improve data quality. Does it make sense? Best, Oleg

By: Cam Bickel

Cam Bickel — Tue, 08 Dec 2009 19:43:34 +0000

Data quality can be improved significantly by adopting PLM but you must be careful not to delay benefiting from PLM by trying to fix legacy data you want to load into PLM. Decide what must be done before you go live and what can be done as you go or not at all.

PLM helps most with data quality issues that arise internally. This is most true for systems that are integrated with PLM. Systems, even internal ones, that are not integrated will still have problems.

What is most challenging is responding to external events. The biggest example is sourcing of purchased parts. Suppliers s change names or go out of business. Products go obsolete or are superseded. There are never enough resources to keep up with this manually, but the automated solutions are expensive and complex.

By: olegshilovitsky

olegshilovitsky — Tue, 08 Dec 2009 15:46:07 +0000

Balu, Agree with you. Well implemented PLM can organize data and as a result Data Quality will be improved. Thank you for your example. Best, Oleg

By: K Balasubramanian

K Balasubramanian — Tue, 08 Dec 2009 12:26:44 +0000

Oleg
PLM has greatly helped in improving Data Quality (DQ) in implementations I have seen. From a state of chaos, duplication and redundency, information has been organized into a predefined, controlled and managed environment and PLM provides a perfect platform to do this.

DQ is directly related to extent of meticulous planning and preparation in early part of PLM implementation.

As companies expand globally, or do acquisition, or expand the foot print of PLM coverage witin the organization DQ is definitely has become an important factor to take care.

rgs
Balu

By: olegshilovitsky

olegshilovitsky — Sun, 06 Dec 2009 23:31:57 +0000

Feyzi, I agree with you. Business and PLM should go together, but it not always happens and PLM often by mistake belongs to pure engineering software. However, I can see changes that happens, so in my view, the trend is positive. Best, Oleg

By: Feyzi Bagirov

Feyzi Bagirov — Sat, 05 Dec 2009 15:24:04 +0000

The average corporate dw is going to be 5 PB very soon. Some companies (eBay) already there. I am not a PLM user, but it would make sense to have adedicated team/department for data management. This trend takes place in many organizations (including DS) on the business side, I don’t see why engineering data management is different. The question could be is should there be one team or dept dedicated to both business and PLM data management?