From aerospace to automotive to consumer electronics, industries are in a race to develop next-generation materials that are lighter, stronger, more sustainable, and more efficient. The problem is, traditional materials discovery is a long, slow, trial-and-error process hampered by limited computational accuracy, high R&D costs and high failure rates. That’s because, if researchers want to discover new materials to, say, make tires on heavy electric vehicles last longer, they must first conduct years of research, develop numerous prototypes, and physically test each one for different criteria, in different physical environments, over long periods of time.
Recent advances in AI and Large Language Models (LLMs) have helped researchers accelerate materials science R&D by enhancing data analysis and extracting insights from existing literature. However, to achieve new materials science breakthroughs, researchers need higher levels of speed, precision, and computational chemistry capabilities that LLMs cannot provide.
Large Quantitative Models (LQMs) represent the next evolution in AI-powered materials science. What makes them so unique and powerful is that they incorporate fundamental quantum equations governing physics, chemistry and biology, so they intrinsically understand how molecules behave and interact. When paired with generative chemistry applications, LQMs can search the entire known chemical space to design molecules with specific desired properties. LQMs can also power quantitative AI simulations, which enable researchers to virtually test how molecules or compounds behave in certain environments – billions of times – before building physical prototypes and/or conducting lab experiments on the most promising candidates. A byproduct of these simulations is a treasure trove of highly accurate synthetic data that can be used to further train the LQMs, making them faster, smarter and more effective.
This convergence of AI and quantum equations enables materials science researchers and manufacturers to:
For manufacturing leaders, the ability to integrate LQMs into their workflows means staying ahead of global competitors, potentially reducing supply chain challenges, and unlocking entirely new products and design possibilities.
There are so many promising opportunities for LQMs to transform industries, including aerospace, agriculture, automotive, chemicals, construction, defense, energy, environmental, manufacturing, and more. Here are just four examples of what’s currently underway (and what we may expect in the future).
1. LQM-Driven Alloy Discovery for Defense and Transportation
SandboxAQ, in collaboration with the U.S. Army Futures Command Ground Vehicle Systems Center, is revolutionizing alloy development with AI-driven Integrated Computational Materials Engineering (ICME 2.0). Using machine learning and high-throughput virtual screening, the project identified five top-performing alloys from over 7,000 compositions, achieving 15% weight reduction while maintaining high strength (830-1520 MPa) and elongation (>10%) while minimizing the use of conflict minerals like tungsten, cobalt, and nickel. This innovation accelerates alloy discovery and enhances the performance of next-generation armored vehicles.
2. Revolutionizing Battery Lifespan Prediction with LQMs
SandboxAQ’s LQMs achieved a breakthrough in lithium-ion battery research, reducing end-of-life (EOL) prediction time by 95% while delivering 35x greater accuracy with 50x less data by predicting EOL with a mean absolute error (MAE) of just 11 cycles using 40 cycles of UHPC data. Now trained on over 1 million hours of NOVONIX’s Ultra-High Precision Coulometry (UHPC) cycle data and achieving predictions with as few as 6 cycles, LQMs cut cell testing from months to days, potentially saving manufacturers millions in R&D. This innovation accelerates battery development by up to four years, driving faster adoption of next-generation energy storage solutions.
3. LQM-Powered Catalyst Design for Sustainable Chemistry
SandboxAQ, DIC, and AWS are revolutionizing catalyst design with LQMs. Using QEMIST Cloud and high-performance computing, the team applied the iFCI method to accurately predict catalytic activity, uncovering superior nickel-based catalysts previously undetectable with conventional methods. A major breakthrough reduced computation time from six months to just five hours, accelerating the discovery of efficient, non-toxic, and cost-effective catalysts for industrial applications.
4. LQM-Enabled Innovation for Cleaner Energy
SandboxAQ’s partnership with Aramco is transforming materials and chemical process optimization in the energy sector. Using its Large Quantitative Models (LQMs), SandboxAQ is developing a multi-GPU-enabled computational fluid dynamics solver to enhance material design and process efficiency in oil and gas facilities. This approach accelerates innovation, reduces costs, and supports Aramco’s goal of lowering its carbon footprint by optimizing chemical production and downstream processes, driving sustainability in industrial operations.
Despite the promise of AI, manufacturing executives often raise three key concerns: cost, scalability, and the expertise required to implement these solutions. While implementing AI solutions requires upfront investment, SandboxAQ’s quantum chemistry platforms significantly reduce R&D costs by accelerating materials discovery and minimizing time-consuming and costly lab experimentation. Companies leveraging these technologies have already seen substantial efficiency gains and innovative product development breakthroughs.
SandboxAQ’s cloud-native AQChemSim platform ensures that manufacturers of all sizes can access quantum-accurate simulations with current high-performance compute infrastructure. This democratization of materials discovery enables rapid adoption across industries.
The demand for quantum-AI expertise continues to grow, but SandboxAQ’s platforms simplify adoption by integrating AI-driven materials discovery tools that require minimal specialized knowledge. This enables companies to harness quantum-powered insights without the need for dedicated quantum computing teams.
LLMs excel at processing and extracting insights from existing text, making them valuable for optimizing workflows and knowledge management. However, they fall short when it comes to tasks like drug or materials discovery, which require a deep understanding of molecular interactions and the underlying physical laws.
In contrast, SandboxAQ’s LQMs are purpose-built for scientific discovery. Trained on the fundamentals of physics, chemistry, and the governing equations of molecular interactions, LQMs are capable of exploring vast chemical spaces to design entirely new molecules. Unlike LLMs, which rely on textual data, LQMs leverage quantum-accurate simulations to predict chemical properties with orders-of-magnitude greater accuracy.
This allows LQMs to optimize a wide range of chemical parameters—such as strength, weight, stability, cost, and sustainability—simultaneously. By replacing slow, trial-and-error experimentation with fast, multi-dimensional search, LQMs dramatically accelerate the discovery and design of materials and drugs. This makes them uniquely suited to solving real-world business challenges involving strategically important materials.
SandboxAQ’s LQMs are unlocking once-impossible breakthroughs, transforming how industries develop next-generation materials. Whether it’s revolutionizing battery lifetimes, accelerating semiconductor advancements, or optimizing alloys for defense, our AI-powered solutions are setting new industry standards.
At SandboxAQ, we are committed to partnering with industry leaders to unlock the full potential of LQMs, from next-gen battery technology to quantum-enhanced catalysts.
