An international research team led by Flinders University in collaboration with Khalifa University in the United Arab Emirates has developed an advanced machine-learning platform designed to accelerate the discovery of next-generation computer chips and electronic materials. The study, published in ACS Materials Letters, introduces a smart materials discovery engine capable of drastically reducing the time and cost associated with traditional laboratory and computer simulation methods used to identify new semiconductors. The primary challenge in material science is the vast number of possible chemical combinations. Testing these options sequentially through conventional means is prohibitively slow and expensive. To address this, the team created an AI system that learns the underlying chemical rules governing gallium-based materials. Unlike random search methods, the platform predicts entirely new material compositions with specific desired electronic properties. The system was trained on thousands of existing semiconductor materials from international databases and utilizes Bayesian optimization to intelligently navigate the search space. This approach continuously identifies promising candidates while filtering out chemically impossible combinations. A critical feature of the AI is its ability to validate chemical realism and physical stability before recommending materials. This rigorous pre-screening prevents wasted effort and significantly shortens the timeline for experimental validation. The study successfully generated multiple new gallium-based semiconductor candidates that were not present in current databases. Gallium, a critical mineral sourced in Australia, is already widely used in electronics. Gallium arsenide, in particular, is essential for microwave circuits, high-speed switching circuits, and infrared applications. The research focused on targeting specific band gaps, a property that dictates how a semiconductor interacts with electricity and light. Different band gaps are required for various technologies: smaller gaps are ideal for solar energy harvesting, medium gaps are crucial for LEDs and optical devices, and larger gaps are necessary for high-power electronics and radiation-resistant systems. By precisely tuning these properties, the AI provides a tailored pathway for developing materials suited for wearable electronics, communication systems, smartphones, medical devices, and solar panels. Lead author Associate Professor Vi-Khanh Truong from the Flinders College of Medicine and Public Health Biomedical Nanoengineering Laboratory emphasized the efficiency of the new method. The platform does not merely generate random formulas but ensures that proposed materials are viable before they reach the testing phase. This innovation marks a significant step forward in optimizing the development cycle for future high-tech applications, leveraging artificial intelligence to solve complex chemical problems that have previously hindered progress in semiconductor technology.