Our References in Variation Management
Fraunhofer Institute for Experimental Software Engineering IESE
Our product line approach (PuLSETM) brings a positive return on investment already after the release of two to three products.
With the help of PuLSETM, one of our customers was able to release five times more products, which were four times more complex, and had three times more features, without increasing the effort.
The adoption of our variation management approaches by customers has allowed them to release new product variants up to twenty times faster.
New Markets and Innovation
Our scoping workshop and further product portfolio planning services have made it possible for several customers to quickly enter a new market.
In one use case, our innovation and evolution plan approach improved the initial product portfolio plan by adding 21 new features and changing the description of 11 features. These improvements were already implemented in the very next version of the product variants.
Reuse and Maintainability
By using PuLSETM, an IESE customer could increase the reuse rate from 17% to 50% and reduce architectural divergences from 17% to 1%, thereby improving maintainability.
In general, the adoption of our variation management approaches has allowed our customers to increase their reuse rate by up to 88% and reduce module dependencies by up to 56%.
Resource consumption is normally a concern for embedded systems. In restructuring families of products by means of reference architectures, we achieved the intended reuse and productivity goals with no significant increase in resource consumption (below 3%).
Our architecture engineering approach for variation-rich software systems integrates flexibility concepts that facilitate customization.
In a customer’s product line, 25% of the existing code contained unused duplicated code, which could be removed using IESE’s dead code detection techniques.
Using one of our tool-based analyses, we analyzed within 15 minutes a product portfolio consisting of four products, each having 1.5 million lines of source code. In the end, we confirmed that over 50% of the code could be transformed into reusable assets, and we listed all the source code components that could be made reusable.
For each source code part of 14 product variants, we correctly identified all the locations where the code was used in the other product variants. Our customer then knows all the locations to be corrected if a bug is found in a specific product.
More information can be found in these selected publications: