MIT Researchers use Machine Learning to Expedite Research on New Battery Materials

With the rise in awareness of the harmful effects of fossil fuels and the increase in sustainable energy and Electrical Vehicles, we have seen a massive shift in the automotive industry. Still, old batteries have been a factor that has slowed the industry down by quite a bit. This research started to pace up on the batteries on how to increase the energy storage capacity and different materials that can work. And now, a possible breakthrough has been found by MIT researchers with the help of machine learning-based tools that can create computer simulations quickly and effectively, as creating computer simulations is a very long and extensive calculation process.

With this system, scientists will have a standardized approach to building models that can correctly compute the trial and error process with different materials. These models are then used to create computer simulations where the scientists can see how the materials interact from the microscopic to the macroscopic level. With the details in these simulations, scientists can see why certain materials do better than other materials, and while the well-studied material, the computer simulation is more accessible. There are pre-built models for them, but they still have errors and require additional tweaking. This model reduces the time taken by the model to build simulations and eliminates any physical fine-tuning needed for the simulation. This model has come at a reasonable time as scientists are looking into a new sort of battery known as solid-state batteries. The challenge comes when building big enough versions of these batteries for EVs because of how the chemistry works since ions don’t like to flow in solids but prefer liquids because of higher intermolecular distance.

No doubt, this method developed by a single researcher is a great help to the battery industry. There is hope for this to be better in the future as to how much progress the machine learning industry is making in a short time. With better techniques and software, this method will become even more critical than it is right now. But for now, we should be expecting this method to get more exposure as more and more batteries and EV manufacturers realize the potential of this product.

Reference: https://news.mit.edu/2022/pablo-leon-battery-materials-0824

Please Don't Forget To Join Our ML Subreddit