Product planning and development in companies must become more efficient while maintaining or increasing the ability to innovate. The use of data from previous products and other references offers significant potential for increasing efficiency. One approach to data-driven engineering is SGE - System Generation Engineering. SGE involves utilizing suitable internal and external references to develop subsystems with carryover variation where possible to reduce costs and risks. To leverage innovation potential new development through attribute and principle variation is applied where necessary. Cross-generational SGE modeling combined with contextual and market success data generates data sets that describe how system generations evolve. With an evolutionary perspective on product engineering and innovation, this study established a research dataset with 14 literature based, retrospective case studies on the evolution of technical systems and derived insights into the evolution of products from a cross-case analysis to support product engineering through heuristics and evolution patterns.