Can AI make Li-ion Batteries Hotter in the Market?

Can AI make Li-ion Batteries Hotter in the Market?

By Energy CIO Insights | Thursday, August 29, 2019

A paradigm shift is predicted for the battery technology especially for Electric vehicles. AI holds the key to unlock the news features and improvements in the batteries.

FREMONT, CA: In a world run by riches, cash, and technology, diamonds are good until they sparkle; currency is valid until it’s in rotation, and technology is useful until the battery is fully charged. In an energy-hungry world like the present, the demand for nickel in the market has risen by 30 percent in the last two months widening the gap as the supply stagnates and demand accelerates.

For small appliances such as cell-phones with batteries, or rechargeable batteries, the big leagues of today, the battery packs which power electric cars such as the Tesla, Prius, and Leaf are powered by lithium-ion batteries. From being a metal loosely called a one-trick pony, heartfelt gratitude is contributed to its dominant end-use in cells. Nickel has suddenly become a two-trick pony, and if electric cars become the next big thing as predicted, then a shortage in the years to come is definite.

At present, no manufacturer has the battery technology to claim that an EV battery takes equal time to charge up as a tank full of gas. The only option available at present is the Tesla Model S 100D 335 miles, which is said to reach a full charge in 75 minutes. Another product option is from the SP group, the EV network in Singapore, which claims to take only 30 minutes.

Although Lithium-ion battery can solve many problems, symmetric to its applications, there exist challenges as well. These challenges range from the necessity of higher energy density to safety compliances. The process of solving the battery chemistry issues is never-ending and majorly based on the iteration of the design, experimentation, and systematic trial processes.

Terabytes of data have been accumulating in the research facilities as they collect detailed information on the performance to statistics from thousands of batteries in real-time. As a result of which material combinations for its research is endless. In such circumstances, using traditional analytical methods is exceptionally tedious.

To tackle these issues, the application of data science in battery development is suggested, as it solves complex models. AI and ML can assess the collected information and build an algorithm much quicker than the human brain. AI-based systems have the potential to autonomously learn and improve in time without the extra attention.

 AI has sparked many ideas across multiple industries; its potential applications are broad. The technology’s reach ranges from material design and synthesis to experiment design, fault analysis, and waste minimization. Its impact in the field of battery development should be considered without understating. The technology has the potential to browse through tons of data to define a fundamental relationship with the measured data and the battery parameters. Manufacturers can have the luxury now to test the permutations and combinations of electrolytes, anodes, and cathodes at any time.

The advantages presented by augmenting AI technology to research are numerous as it will enable a better understanding of the batteries. Limitless combinations can be tested in no time, meaning the ultimate formulation of materials to build the perfect battery can be found much earlier than predicted. Integration of technology will drastically reduce the necessity of conducting innumerable experiments, cutting down the time as well as the costs for development. For better understanding, a team of 50 researchers experimenting on a particular combination for the battery can save up to $1m in research per month by deploying the capacities of ML.

An initial foray of a battery magnate into a similar technique has achieved extra-ordinary results. The team focused on developing the ultra-fast charging battery technology augmented ML to the research. The innovation of the first of its generation, battery utilized ML to discover that simple tweaks in the construction could result in doubling the number of cycles of the cell. This discovery has inspired the entire industry to consider augmentation of technology with their research projects.

Break-through found by this technique is now being applied in the latest generation of EV batteries. The industry has found out that not one or two, but several other issues need to be overcome to create the ultra-fast charging batteries. Although the breakthrough in the field made a difference, there is yet a long way to go. Hence, the power of data science in innovative forms is combined with AI and expertise in electrochemistry to bring about a climax to complex puzzles presented by battery research.

ML is not the only way in which AI can be utilized in battery development. An interesting way to approach this phenomenon is by applying the technology within a vehicle’s operational software. By doing so, continuous monitoring of battery performance and health, communication of reports for improvement of the product is possible. An essential offering of this system is predicted to be the creation of smart batteries with embedded sensors. These sensors will project self-healing functions, a holistic battery-manager, which can disseminate total awareness of the battery status and modules when needed.

AI has helped innovators transform more than one component simultaneously and has assisted in analyzing the evidence quicker. The integration of technology has fuelled the process and has kick-started the innovation in a way that traditional statistical analysis could never accomplish. The adoption of Batteries for EVs is of paramount importance at present. By bringing down the charging times with technology, the integrated process would not only have presented a cake, but it would be serving with a cherry on top.

Weekly Brief

New Editions