TU Delft Embraces Research Failures in Groundbreaking AI Symposium
Delft, Thursday, 14 May 2026.
TU Delft’s AI Initiative is hosting ‘Bloopers of Brilliance’ on May 27, 2026, a unique symposium where PhD and postdoc researchers celebrate and learn from scientific failures. Breaking academic tradition of showcasing only successes, this event encourages early-career AI researchers to openly discuss setbacks and mistakes. The symposium features poster presentations on failed experiments, panel discussions on research disappointments, and networking opportunities designed to normalize failure as essential to innovation. This pioneering approach reflects TU Delft’s commitment to creating supportive research environments where learning from mistakes drives scientific progress forward.
Event Structure and Academic Innovation
The AI PhD & Postdoc Spring Symposium 2026, officially titled ‘Bloopers of Brilliance: When Science Goes Sideways,’ is scheduled for May 27, 2026, from 13:00 to 17:30 at YES!Delft [1]. Organized by the AI PhD Committee in collaboration with the TU Delft AI Initiative, the symposium will feature poster pitches, a panel discussion on scientific failures, keynotes from final year PhD students, and networking opportunities [1]. This innovative format represents a departure from traditional academic conferences that typically emphasize successful outcomes and breakthrough discoveries [GPT]. The event specifically targets PhD and postdoc researchers at TU Delft working across all areas of AI-related topics, creating an inclusive platform for cross-disciplinary learning [1].
Participation Framework and Research Themes
The symposium employs a comprehensive participation structure with multiple deadlines to accommodate varying levels of engagement. The final deadline for participating with a poster and abstract was May 21, 2026, while researchers can still register to attend without presenting until May 26, 2026 [1]. The poster market encompasses diverse AI research themes including machine learning and foundational AI techniques, human-centered AI systems, AI applications, fairness and ethical considerations, education and AI, design with AI, and reflexive and critical research on AI [1]. To incentivize participation and recognize valuable failure insights, organizers offer prizes for the best poster and ‘failure abstracts,’ including a three-month subscription to X (formerly Twitter) [1]. This structure demonstrates TU Delft’s commitment to making failure discussions accessible while maintaining academic rigor.
Distinguished Speakers and Expertise
The symposium features a carefully curated lineup of speakers representing diverse AI research perspectives and career stages. Confirmed speakers include Anne Poot, Surya Manoj Sanu, Casper van Engelenburg, Sofia Kotti, Elvire Landstra (postdoctoral researcher in Epidemiology at Tilburg University), Agostino Nickl (Assistant Professor of Architectural Intelligence at TU Delft), and Nazli Cila (Associate Professor in the Department of Human-Centered Design at the Industrial Design Engineering Faculty) [1]. Nickl brings international experience as a former Visiting Professor at Politecnico di Milano, Senior Teaching Fellow at UCL, and instructor at prestigious institutions including ETH Zurich, the Bartlett School of Architecture, and Oxford School of Architecture, along with industry experience at Arup’s Digital Studio in London [1]. Cila co-directs the AI DeMoS Lab, while other faculty members like Megha Khosla, an Assistant Professor in the Multimedia Computing Group within the Intelligent Systems Department, focus on developing explainable and privacy-preserving machine learning methods and recently received the NWO Vidi grant [1].
Broader AI Research Ecosystem at TU Delft
This symposium occurs within TU Delft’s expanding AI research infrastructure, which includes over 50 AI labs coordinated by the TU Delft AI Initiative to facilitate interdisciplinary collaboration across eight faculties [5]. The university’s commitment to AI innovation extends beyond academic research to practical applications, as demonstrated by projects like the City AI Lab’s work on urban design, where researchers analyze millions of street-level images and gather input from over 3,000 Dutch residents to understand how artificial intelligence can help cities design more attractive neighborhoods [5]. Additionally, the emergence of student-led initiatives like the Delft AI Hub, founded by seven students from different faculties, reflects the growing grassroots interest in AI across the university community [2][4]. The Hub serves as a platform for collaborative projects, workshops, and networking, with founders emphasizing accessibility for TU members at all experience levels, from complete beginners to advanced practitioners [2][4]. This multi-layered approach to AI education and research, spanning from formal symposiums addressing research failures to informal student networks promoting collaboration, positions TU Delft as a comprehensive hub for AI innovation in the Netherlands.