Semantic descriptions make processes, data and knowledge from chemical syntheses machine-actionable in autonomous laboratories.
Source: BAM
The development of new and optiumized materials is increasingly driven by data, interlinked knowledge and automation. In so‑called self‑driving laboratories, robotic systems perform chemical syntheses autonomously, analyze results and optimize processes without continuous human intervention. For this interaction between robotics, software and artificial intelligence to work reliably, experimental procedures must be described in a precise, complete and machine‑actionable way. This is exactly where the presented WeChemSyn Ontology comes into play.
In chemistry and materials science, synthesis procedures are traditionally documented as free text. While this format is easy for humans to read and understand, it often relies on implicit knowledge and vague expressions such as “overnight”, “room temperature” or “slow addition”. For automated laboratory systems, such descriptions are insufficient, as machines require clearly defined steps and parameters.
The WeChemSyn Ontology provides a structured semantic knowledge representation for wet‑chemical syntheses. It supports the systematic capture of recurring process steps, materials, devices and measured quantities in a form that computers can interpret, compare and reuse. The ontology builds on internationally established semantic frameworks developed for materials science and engineering and integrates existing standards.
Using real examples from a self-driving laboratory at BAM, it is demonstrated how chemical syntheses can be represented in knowledge graphs. These graphs explicitly link individual process steps and make both sequential and parallel operations transparent. As a result, experiments become easier to reproduce, to share between laboratories and to evaluate systematically.
In the long term, this approach lays the foundation for making chemical syntheses transferable and executable across different automated laboratories worldwide. The work therefore represents an important contribution to the digital transformation of materials research and to the accelerated discovery of new advanced materials.
WeChemSynOntology: semantic modeling of wet chemical syntheses in a self-driving lab for nano- and advanced materials
M. Schilling, H. Bresch, B. Bayerlein, B. Ruehle,
Digital Discovery, 2026