Scientific research is a dirty business. The path to learning new things and making discoveries is paved with hard work, hard thinking, and many dead ends. It’s a time-consuming, expensive endeavor, and for every success there are thousands upon thousands of failures.
It’s a process so inefficient you’d think someone had already automated it. The concept of the self-driving laboratory aims to do just that and could revolutionize materials research in particular.
Leave it to the Auto-Lab
Materials research is a complex field and challenging to work in. A lot of the work is finding improvements to existing materials to make them harder, better, faster, or stronger. There is also the possibility of discovering entirely new materials with unique properties and abilities beyond those we already know.
In the modern world, simply going outside and digging up a new type of rock or finding a new type of tree is not enough. All low-hanging fruits are already gone. Materials research today requires a sound understanding of physics and chemistry. It’s about figuring out how to use those principles to make something better than what we’ve seen before.
This is where artificial intelligence and computers come into play. Rules that we discovered in chemistry and physics can be programmed into an intelligent system. It’s then an easy jump to have this system apply these rules in different ways to optimize the desired results. For example, an AI system can be asked to synthesize a specific chemical in the most efficient way given a specific set of precursor chemicals. Then the AI can run through all the possibilities and determine the best course of action.
However, the concept becomes most compelling when an AI system is given the ability to conduct its own experiments in the real world. Laboratory automation has advanced to the point where robots can perform experiments much faster and more efficiently than human scientists could do by hand. Give the AI the hardware to run experiments and measure the results of its work, and it can then use the results to direct further experiments toward its predetermined research goals. Congratulations – you’ve built a self-driving laboratory!
In practice
Far from a mere theory, researchers around the world are already building self-driving laboratory systems. One of the best known is the so-called Artificial Chemist, developed by researchers at the University of Buffalo and North Carolina State University. The aim of the project was to develop an automated system for conducting chemical research for the discovery and development of commercially desirable materials.
It was designed to conduct chemical research on materials that can be made using liquid solutions. The system is tasked with finding a way to synthesize a material that satisfies a set of desired parameters and is conducting experiments of its own to determine how this can be achieved. When testing the system, the task was set to synthesize quantum dots with various desired parameters. Through experimentation, Artificial Chemist was able to find ideal techniques for making the dots, including identifying the correct chemical precursors.
Far from a simple computer simulation, Artificial Chemist performs real chemistry itself and measures the results. The system was equipped with fully autonomous chemical reactors. They are also designed to stay clean without picking up chemical residues that would interfere with experiments. The system can mix chemicals and perform an entire chemical synthesis all by itself.
The system was developed with research and manufacturing in mind. It can be commissioned to make quantum dots for a specific wavelength of light and will initially spend time on research experiments to determine the best way to make them. Once this process is complete, typically 1-10 hours, the system can start mass-producing the points.
research+
Overall, however, the basic principle can be applied to all types of research processes. All you have to do is give an appropriate AI system the means to experiment and the means to verify the results of your work. It can then take the logical steps to continue its work towards its predetermined research goals.
The advantages of such systems are manifold. Where parts of experiments may previously have been automated by robots, self-driving labs go even further. They enable scientists to set a goal and the automated laboratory works towards a solution completely independently. This allows research to be carried out with less labor and manpower, with progress being made much faster and much cheaper than before. Also, the ability for quick calculations and experiments can allow an AI to quickly run tests to combine regular ingredients in unexpected ways, leading to surprising unconventional results. Some researchers expect these systems to offer a 10-fold cost and time advantage where goals that previously took 10 years and $10 million were achieved in a year and for just $1 million.
Of course, such systems will not render human researchers obsolete. Creativity is hugely important in science and technology and has led to some of our greatest advances. For example, an AI could be tasked with making stronger and lighter metal alloys. However, given these man-made prejudices, it would never reach the brilliance of composite materials like carbon fiber.
A great follow-up is image synthesis AIs, which have skyrocketed in popularity this year. Initially, the exaggeration said that artists and photographers were out of work and human endeavors in this field were over. Then, weeks later, it turned out that this was just a new type of tool to be wielded and used by people best at exploiting them.
These “autonomous labs” are likely to become important tools in industrial research and development labs doing everything from developing new materials to uncovering new molecules of potential medical interest. Talented researchers will work to make the most of the robotic resources at their disposal and ensure they are used in the most effective manner towards their broader research goals in general. With much of the research offloaded to the robots, human scientists have more time to think about the big picture.
Banner image: © xiaolangge / Adobe Stock.
