Autonomy in materials research: a case study in carbon nanotube growth

Abstract

Advances in materials are an important contributor to our technological progress, and yet the process of materials discovery and development itself is slow. Our current research process is human-centred, where human researchers design, conduct, analyse and interpret experiments, and then decide what to do next. We have built an Autonomous Research System (ARES)-an autonomous research robot capable of first-of-its-kind closed-loop iterative materials experimentation. ARES exploits advances in autonomous robotics, artificial intelligence, data sciences, and high-throughput and in situ techniques, and is able to design, execute and analyse its own experiments orders of magnitude faster than current research methods. We applied ARES to study the synthesis of singlewalled carbon nanotubes, and show that it successfully learned to grow them at targeted growth rates. ARES has broad implications for the future roles of humans and autonomous research robots, and for human-machine partnering. We believe autonomous research robots like ARES constitute a disruptive advance in our ability to understand and develop complex materials at an unprecedented rate.

Printable Materials for the Realization of High Performance RF Components: Challenges and Opportunities

“…We also wish to highlight that rapid advancements are needed in the design of experiments necessary to address the challenges we have outlined. Here, machine learning and autonomous experimentation could be integrated into the current research field as they have shown great promise for carbon nanotube synthesis [131]. Ultimately, we envision processes associated with AM to be highly automated during the full-scale production workflow with robotics, especially when needed to go between instruments to print at different scales (nm to cm) and final system assembly.…”

A machine learning approach to thermal conductivity modeling: A case study on irradiated uranium-molybdenum nuclear fuels

“…Machine learning methods including, but not limited to, deep learning have recently been applied to more challenges in molecular and materials science fields [21]. Applications of such methods include the following: development of accelerated materials design and property prediction [22,23,24,25,26,27,28,29], process optimization [30], discovery of structure-property relationships [31,32,33], characterization of structure and property data [34], and image classification and analysis [34,35,36,37,38,39]. Such applications span multiple lengthscales (macro-to nano-scale) and a variety of material systems (inorganic oxides, electrolytes, polymers, and metals) [40].…”

Molecular characterization of macroscopic aerogels of single-walled carbon nanotubes

“…We envisage the characterization protocol to follow this sequence: i) determination of SWCNT family distribution by Raman using 3-4 laser lines (resolution of ±50 meV); if the distribution is narrow, for example, less than ≈ 3 families observed per laser energy, then ii) perform HRTEM electron diffraction and diameter measurements (requiring resolving diameter and chiral angle differences of < 0.1 nm and 1 • respectively) and iii) carry out absorption and emission measurements (example of optical absorption spectrum for an aerogel ESI). An interesting possibility is to combine these tools with high-throughput synthesis and analysis methods [32].…”