Every year, software engineering research produces innovative tools, visualizations, analyses, and more recently AI-powered assistants. Yet many of these innovations never make it into the daily work of developers.

Based on real-world experiences from introducing research outcomes into a large industrial software ecosystem, this keynote explores what happens after the paper is published: the difficult journey toward adoption. We will discuss successes, failures, unexpected barriers, and the lessons learned while deploying research-driven solutions in environments characterized by legacy systems, long product lifecycles, and strong operational constraints.

From model-driven engineering and software visualization to automated test generation and AI-assisted development, the talk highlights a simple observation: the most advanced solution is not always the one that succeeds. Impact comes when research becomes part of the developer’s natural workflow. Ultimately, this keynote is an invitation to rethink how we measure success in software engineering research, by integrating real-world adoption.

Benoit Verhaeghe holds a Ph.D. in Computer Science and works as a research leader at Berger-Levrault. His collaboration focuses on creating innovative solutions for migrating and enhancing software systems, leveraging meta-models. Benoit is also active on GitHub, where he shares his public projects. Beyond his technical work, he has authored several scientific papers, and blog posts.

For more insights into his research, you can visit his personal website. Feel free to explore his work further and follow his exciting contributions in the field of computer science!