In an era of fully-self-driving cars and electric vehicles, one might think the automotive industry has reached the pinnacle of technological innovation – at least until cars start flying! Over the past five years, the integration of AI and more powerful processing capabilities into automobiles has brought forth such safety innovations as adaptive cruise control, lane keeping, and automatic emergency braking; enhanced vehicle monitoring and information systems; more personalized and interactive driver experiences; and increasingly more sophisticated automation capabilities. As both AI and processors become faster and more powerful, we can expect to see similar automotive innovations in the future. But what about AI innovations outside the vehicle?
Today, we stand at the threshold of an automotive revolution driven not by a new type of vehicle or eyebrow-raising features, but by advanced simulation technologies and materials science capabilities powered by a new kind of AI – Large Quantitative Models (LQMs). These innovations will help improve automobile performance, safety, and sustainability, optimizing everything from design to final delivery.
Throughout history, automotive progress has been shaped by scientific advancements in physics, chemistry, and thermodynamics. Every material, part, system and vehicle design undergoes exhaustive testing by engineers – in the lab and on the test track – to ensure its safety, reliability, and optimal performance. Advances in computing power, computer-aided design, and AI have accelerated automotive R&D, but every new innovation still requires extensive trial-and-error testing in the physical world. With limited facilities available to conduct such tests, automotive R&D cycles are costly and time-consuming – leading to increased costs for consumers and businesses. LQMs can help mitigate these obstacles by simulating real-world designs, material properties and driving conditions digitally, reducing development timelines from years to months.
While Large Language Models (LLMs) continue to grab headlines, their applications in the automotive sector remain limited to customer support, marketing, and image generation. LQMs, on the other hand, are purpose-built for processing scientific equations and real-world data. Unlike traditional simulation methods, LQMs handle vast datasets and perform complex calculations to drive tangible advancements in:
Separate but related, LQMs are also being used to strengthen cybersecurity, which could eventually extend to vehicular systems. They are also being tested by the aviation industry as an alternative navigation system in the event of GPS signal loss, which could expand to land-based use cases as well. Major energy companies are leveraging LQMs to explore the creation of biofuels or stable hydrogen, which will help reduce the 6 billion metric tons of carbon dioxide produced annually by road transportation worldwide.
Advanced materials are key to modern automotive engineering, influencing everything from fuel efficiency to safety. For example, research shows that a 10% reduction in vehicle weight can improve fuel economy by 6–8%, while shedding 100 kg of weight can cut CO₂ emissions by nine grams per kilometer.
SandboxAQ’s proprietary LQMs enable researchers and engineers to explore the vast chemical space and discover novel material compositions that meet specific requirements, and then simulate them millions of times – factoring in real-world conditions – to test their properties and performance. This significantly accelerates R&D and lowers costs for creating advanced materials, composites and alloys for lighter, safer, more fuel efficient vehicles.
Putting this into practice, the U.S. Army is currently leveraging SandboxAQ’s AQChemSim platform to help discover advanced alloys for its armored vehicles. Using LQMs, SandboxAQ helped the Army identify five top-performing alloys from over 7,000 compositions, achieving 15% weight reduction while maintaining high strength (830-1520 MPa) and elongation (>10%) and minimizing the use of “conflict minerals” like tungsten, cobalt, and nickel in its armor. This same approach is now being adopted by leading automobile manufacturers to drive similar improvements.
The surging worldwide demand for EVs has led to increased global demand for advanced batteries – and increased pressure on global supply chains. Rapidly designing and producing better batteries – while simultaneously navigating geopolitical sensitivities around cobalt and nickel, and geological limitations on critical materials such as lithium – remains a major challenge for the EV industry.
For example, traditional testing methods for lithium-ion (Li-ion) batteries (LIBs) require months or even years to predict end-of-life (EOL) degradation. However, using 4 million hours of Ultra-High Precision Coulometry (UHPC) battery cycle data from leading battery materials and technology company NOVONIX, SandboxAQ’s LQMs achieved a 95% reduction in testing time for predicting Li-ion battery lifespan with 35x greater accuracy using 50x less data compared to traditional approaches. This breakthrough reduces cell testing times from months to just days, enabling battery manufacturers to focus resources on the most promising designs and cut R&D time and costs considerably.
Similarly, SandboxAQ is working to significantly advance battery shelf-life testing. In collaboration with the U.S. Army Combat Capabilities Development Command (DEVCOM), SandboxAQ compiled a dataset of over 2 million hours of battery testing data, which will help the Army assess the status of the Li-ion batteries used in a wide range of applications. We’re also helping them develop new battery chemistries, materials and designs for diverse applications such as EVs, Unmanned Aerial Vehicles (UAVs), and portable power solutions.
The insights and learnings from these collaborations will help battery manufacturers improve battery performance, vehicle range, safety, fast-charging capabilities, and integration with existing infrastructure while mitigating costs and environmental impact.
For the auto industry, the road ahead is clear – and paved with numerous possibilities unlocked by LQMs. From creating novel materials and fuel sources to improving safety, performance and sustainability, LQMs are not just accelerating innovation, they are redefining what’s possible. Using physics-based quantitative AI models – instead of language-based LLMs – auto manufacturers can significantly reduce R&D time, cost and risk, discover and implement new tech and engineering innovations, and speed new and better products to market. Not only will this create value for their organizations and key stakeholders, it will create enthusiasm and delight among consumers and commercial operators. And who knows: maybe one day LQMs will play a pivotal role in developing flying cars – taking auto manufacturing and design to even greater heights.
To learn more about how LQMs can revolutionize the automotive industry, please contact us today and visit our website.
