Self-Driving Laboratories

Comprehensive review of hardware and software stacks for autonomous chemistry laboratories, including case studies.

Tutorials, workshops, and reading lists from the University of Toronto's centre for self-driving labs.

Practical overview of building blocks for self-driving labs: automation, optimisation, and AI integration.

Aspuru-Guzik group's benchmark for optimisation algorithms on real chemical datasets. Tutorial notebooks for running BO on reaction emulators.

Interactive tool generating BoTorch/Ax code for common experimental design scenarios. Useful for quickly prototyping a BO loop for a real experiment.

Reviews the evolution of autonomous labs from flow chemistry to AI-directed multi-step synthesis.

Coscientist: LLM-orchestrated agent controlling web search, code execution, and a robotic chemistry platform.

Introduces Olympus and benchmarks 10 optimisation algorithms on 10 real chemical reaction datasets.

Integrates ML retrosynthesis planning with automated flow synthesis — one of the first self-driving synthesis platforms.

Benchmark framework for chemical optimisation algorithms with emulators for 10+ real reaction and materials datasets.

BO tool designed for chemical reaction optimisation over categorical and continuous experimental variables.

Open-source robotics software and protocols for self-driving lab infrastructure from the University of Toronto consortium.

10+ real chemical reaction and materials datasets with noise-emulator surrogates for benchmarking optimisation algorithms in self-driving lab scenarios.

42,000+ perovskite solar cell device results with composition, fabrication conditions, and efficiency metrics. A real closed-loop optimisation target for SDL demonstrations.

Thermochemical, spectroscopic, and reaction kinetics data for 80,000+ compounds from NIST. A reference source for validating SDL measurements and models.