AI in the laboratory: from tool to research partner

Artificial intelligence is conquering the laboratory step by step. During WoTS 2026, it became clear that AI is no longer being used solely for analyzing data, but is increasingly becoming part of the research process itself. From medical imaging to self-driving laboratories: the technology is shifting from a supporting tool to an active partner in research and innovation.[1]

In healthcare, the challenge lies not so much with the technology itself, but with its implementation. According to healthcare consultant Jawad Handizi, various AI applications are now available for imaging, diagnostics, and data processing. However, practice shows that successful implementation depends on factors such as local validation, budgets, acceptance by healthcare professionals, and a clear business case.

It is striking that hospitals are primarily interested in solutions that accelerate or simplify daily operations. Doctors and specialists want to know not only whether an AI system can recognize an abnormality, but above all how it contributes to shorter diagnostic pathways, fewer administrative tasks, and more efficient work processes.

Handizi expects that AI will increasingly take over routine tasks. Nevertheless, human expertise remains necessary. AI systems support decision-making but do not replace the healthcare professional. The challenge therefore shifts from technical development to organizational adoption.

Self-driving laboratories

While AI in hospitals primarily supports existing processes, research laboratories are working on the next step: the self-driving laboratory. Within the national Big Chemistry program, universities, research institutes, and companies are developing laboratories where robotics, data analysis, and machine learning come together. The goal is to develop new materials, formulations, and chemical processes faster and more efficiently.

Central to this is the concept of the 'self-driving lab'. Robots conduct experiments, analyze the results, and use machine learning to automatically determine which follow-up experiments yield the most new knowledge. As a result, not every possible experiment is performed anymore. The system independently searches for the most promising route through a large field of research.

This approach not only saves time, but material consumption also decreases significantly because fewer measurements are required to achieve an optimal result. This can make a major difference, especially for complex research questions involving many variables.

From automation to a new research infrastructure

The Big Chemistry consortium is investing not only in technology but also in a new research infrastructure. Shared data, standardized workflows, and reusable automation modules play an important role in this. As a result, researchers can build on previous work more quickly and do not have to reinvent the wheel as often.

In addition, automation can contribute to a persistent problem within science: reproducibility. Standardized, automated workflows reduce the likelihood of differences between laboratories and make research results more comparable.

The next phase

The common thread of the afternoon was clear. AI is moving away from the role of an analytical tool and evolving into a system that actively directs processes. In hospitals, AI helps doctors work faster and more accurately. In laboratories, AI is increasingly determining which experiments should be performed next. This creates a new form of collaboration between humans, machines, and data.

The question for the coming years is no longer whether AI will find a place in laboratories and research organizations. The question is how researchers, technicians, and organizations learn to effectively utilize this new partner.

[1] Seminar – Implementing AI in Healthcare: Challenges, Successes and Practical Insights – FHI, Federation of Technology Branches