You’ve completed a needs assessment (gap analysis), identified and prioritized user needs and functionality. After a few months of planning, design, configuration, verification testing, and coding your new LIMS is ready to deploy. If you are replacing an existing LIMS, there is one last critical decision to make–which go-live option will you use? There are three methods to consider: parallel, incremental, or big bang. Each approach has its own challenges and advantages.
Parallel Testing uses multiple versions or components of software, using the same input on different systems simultaneously to reduce time required. The goal is identical output from each system.
A major benefit is confidence that the new LIMS completely meets the lab’s needs. This comes from the unique ability to run each system in parallel until the team feels confident in the system’s output based on test and user case scenerio results. Once testing is complete the laboratory team can fully validate and sign off on the new LIMS.
The disadvantage of this approach is the cost and technical complexity, –maintaining two environments is clearly far more expensive than one and the integration capabilities (i.e. instruments, enterprise applications) required to maintain parallel coordination of data can be significant.
In some cases, an incremental approach may be the best solution. It reduces the risks associated with big-bang, as focus is on a single phase at a time that deals with discrete parts of the laboratory operation.
An example incremental scenario: Phase I includes getting the LIMS functioning (system configuration) from setting up calculations, hold times, preparation steps to tracking samples and automatically email reports and invoices. Phase II includes instrument integration and all of the associated verification testing and validation that must occur before going live. Phase III likely consists of enterprise integration with systems such as SAP, GIS, SCADA, or manufacturing solution or accounting packages.
Challenges with an incremental launch strategy are related to the complexity of the LIMS (configuration and feature/function breakdown), data, and the laboratory’s business processes.
The big bang data migration concept is simple–build the data migration architecture and then move the data to the new LIMS and environment in one process. Typically, during these types of deployments, people are working around the clock. This approach is more common for less complex IT systems, and rarely the preferred method for LIMS. These are typically used in emergency migrations, when the current system is abruptly unavailable and a new system must quickly be deployed. All migration tasks are greatly compressed, and any issues that are uncovered that did not appear during testing must be immediately resolved.
Benefits of the big bang approach include: quick archival of the legacy system and timely realization of Return on Investment, no need for data migration from the historical system to the new system, and no synchronization issues. Challenges to the big bang approach include limited capacity for non-mission critical work, limited timeframe that could affect customer service, and difficulty “rolling back” to fix issues
A complete parallet test plan will ensure an efficient execution of your LIMS migration. In deciding which option is best, you must determine your goals and assess the resources required. A LIMS professional can help.
Submit this Form to Subscribe To Our Blog.