Deploying Data Management Applications

Discussion created by AndyFreebody on Nov 26, 2008

When considering the deployment of a data management system, be it for library, component information or design data management, the temptation is to try and achieve all of the possible business benefits of the system in a single step.  This often leads to lengthy implementation programs involving detailed requirements specification, implementation and acceptance testing which invariably result in overly complex customisations that don’t necessarily meet the objectives.  The most productive and successful approach is to deploy the system in phases where each phase is designed to target a specific set of business oriented objectives.  The first phase should always utilise the systems out of the box functionality; this will enable the users to become familiar with the core capability of the tools that they are to use.  This is important because without the familiarity with the tools core capabilities the users cannot be expected to accurately identify the minimum set of customisations that are required to achieve the business objectives.


Applying this to DMS one could consider first deploying the core library management and component information management modules as is by simply configuring the classification and technical parameters so that users can get used to the default use cases that DMS supports.  The next phase could include an interface to the ERP system to automatically load the logistics data into the supply chain model so that the engineers have access to the lead time and cost information.  Then perhaps there could be another phase covering process automation to ensure that the correct process is followed for new component introductions.  In this way each phase would deliver a distinct business benefit and would build on the success of the previous phase while minimising the time to production.