Many in R&D are still held back from achieving their goals. Forty-eight percent of R&D teams report they can't move fast enough to solve revenue-generating challenges in material science, with 52.7% of researchers spending upwards of 8 hours each week on manual research processes. Of those surveyed, 35.2% also state that the insights they are gaining from the R&D processes are not as scalable as they need to be.
AI presents a potential solution to tackle these issues. According to the survey, a significant 69.5% of respondents agree that AI has been instrumental in helping them accelerate discoveries. Material science is a vast and interdisciplinary field that intersects with physics, chemistry, engineering, and biology, focusing on the discovery and design of new materials. Key areas of focus include developing more efficient battery technologies for improved energy storage, creating sustainable and recyclable materials to address environmental concerns, and advancing aerospace materials for lighter and more durable aircraft components.
Despite 55.5% of respondents using AI tools from major providers like ChatGPT and Microsoft Co-pilot, those surveyed have hesitations about its suitability for scientific research. Respondents cited their main concerns as hallucinations (30.3%) and data security (30.8%) for using AI in their R&D efforts.
The survey identifies the top three features that could enhance trust in AI for scientific research: increased data security (51.3%), citations on the origins of data (47.8%), and citations on the quality of data (46.8%).
Anita Schjøll Abildgaard, CEO of Iris.ai, commented on the findings: "Material science is at the cutting edge of scientific research, driving forward new innovations to tackle some of the biggest problems of our time. But with this complexity comes roadblocks in the research process: too many researchers are still mired in manual research processes. There is an opportunity to use AI to put scientific knowledge into the hands of researchers – faster – and accelerate new discoveries.
“There's a pressing need for more tailored, secure, and scalable AI solutions in R&D. To truly meet the demands for scientific innovation, AI tools must be not only powerful but also actionable and trustworthy, enabling researchers to bridge disciplinary gaps and turn insights into revenue-generating outcomes more efficiently."
