Toward autonomous laboratories: Convergence of artificial intelligence and experimental automation

Accelerated Electrosynthesis Development Enabled by High-Throughput Experimentation

“…140 The combination of automation and machine learning has shown considerable promise for speeding up screening by an order of magnitude. 141 A machine learning-based electrolyte metering/mixing/evaluation test stand was designed and assembled by Whitacre et al 142 The test stand was used to query several 2-dimensional electrolyte search spaces discovering novel binary electrolyte solution blends such as aqueous LiNO 3 and NaNO 3 blends with voltage stabilities exceeding 2.8 V. With the help of Bayesian optimization, it took less than a day conduct these searches while conventional manual methods would likely have taken much more time. Implementation of a feedback loop enables autonomous optimization procedures using active learning algorithms.…”

De Novo Design of Molecules Towards Biased Properties via a Deep Generative Framework and Iterative Transfer Learning

“…Such ondemand generation algorithms are necessary for moving toward autonomous laboratories. 25 Among the various DL models proposed recently for molecule design, GANs bring in breakthroughs. Two proof-of-concept GANs, namely ORGAN 26 and ORGANIC, 27 were introduced to generate novel molecules, while the generation is not conditioned on the physicochemical or biological properties with quantitative and continuous labels.…”

What is missing in autonomous discovery: open challenges for the community

“…Scientific experimentation and discovery is teetering on the precipice of a new industrial revolution. Acceleration of science by combining automation and artificial intelligence (AI) has begun to revolutionize the structure of scientific experiments across physics, 1 chemistry, 2–8 materials science, 9–13 and biology. 14 The integration of high-throughput experimentation, AI, data science, and multi-scale modeling have spawned great interest, 15 notable results, 16 and substantive expectations.…”

“…While we will rst describe the current state of the eld, this is by no means a comprehensive review, and we encourage the reader toward reviews and perspectives of SDLs and autonomous experimentation. [1][2][3][4][5][6][7][8][9][10][11][12][13][14]16,[23][24][25][26][27][28][29][30][31][32][33][34][35][36][37][38] Following this, we will turn our attention to barriers and opportunities associated with data, hardware, knowledge generation, scaling, education, and ethics. As the eld of autonomous experimentation grows and SDLs become more common, we hope to see rapid growth in scientic discovery.…”