Advertising

Altrove: Using AI to Predict and Create New Inorganic Materials

In the world of materials development, innovation has been rapidly advancing in recent years. French startup Altrove is poised to make its mark in this field, having already raised €3.7 million in funding. The company aims to address a longstanding challenge in materials research – predicting the existence of new materials. Altrove’s co-founder and CEO, Thibaud Martin, explains that historically, R&D in this area has been slow due to the sheer number of possibilities when combining different chemical elements. However, advancements in artificial intelligence (AI) have enabled teams from companies like DeepMind, Microsoft, Meta, and Orbital Materials to develop models that can overcome these calculation constraints and predict new materials. In fact, Martin notes that in the past nine months alone, more stable materials have been predicted than in the previous 49 years.

While predicting the existence of new materials is a significant achievement, it is only one part of the equation. Altrove recognizes that knowing a material can exist is not enough – the real challenge lies in creating a recipe for its production. Martin emphasizes that a recipe involves not only the elements used but also their proportions, temperature, order, and duration of the manufacturing process. With these factors in mind, Altrove focuses on inorganic materials, particularly rare earth elements. This choice presents a market opportunity as rare earth elements are difficult to source and often come from China, leading many companies to seek alternatives to avoid supply chain uncertainties.

Altrove’s approach involves selecting promising candidates from the pool of predicted materials and using its AI models to generate potential recipes. The company then tests each recipe individually and produces small samples of the resulting materials. To ensure the materials perform as expected, Altrove employs its proprietary characterization technology, which utilizes an X-ray diffractometer. This step is crucial as it allows the company to understand why the produced material may or may not meet the desired specifications.

Altrove’s strength lies in its co-founder and CTO, Joonathan Laulainen, who has a PhD in materials science and specializes in characterization. This expertise enables the company to learn from the characterization process and improve its recipes for new materials. To further enhance this iterative feedback loop, Altrove plans to automate its lab, enabling the testing of multiple recipes simultaneously and speeding up the experimentation process.

Describing itself as a hardware-enabled AI company, Altrove envisions selling licenses for its newly developed materials or partnering with third parties for production. The startup has already secured funding from investors such as Contrarian Ventures, Emblem, and several prominent business angels. Taking inspiration from the success of AI-driven approaches in the biotech industry, Altrove aims to build its automated lab by the end of the year and expects to sell its first asset within 18 months.

With its innovative combination of AI and materials science, Altrove is poised to revolutionize the field of materials development. By overcoming the bottleneck of predicting new materials and streamlining the recipe creation process, the company has the potential to unlock a wealth of possibilities for various industries. As Altrove continues its journey, it will undoubtedly contribute to the ongoing acceleration of innovation in materials research.