Sometimes the most innovative solutions come from the most unexpected places. For Associate Professor Jiangtao Cheng at Virginia Tech, inspiration struck while playing an ancient board game—and that moment of insight could help keep our increasingly connected world running cool and efficient.

Cheng and his research team, including Assistant Professor Zhenhua Tian and Ph.D. candidate Mohammad Shamsodini Lori, have successfully developed groundbreaking AI algorithms that optimize spray-cooling systems by drawing inspiration from Go, a strategic board game invented in China over 2,500 years ago. Their findings, recently published in Artificial Intelligence Review, represent a beautiful marriage between ancient wisdom and modern technology.

The breakthrough addresses a critical challenge facing our technology-dependent society. Every machine that runs on electricity or fuel generates heat, and managing that heat effectively makes the difference between reliable power grids and blackouts, or functioning data centers and widespread outages. Spray cooling—where liquid droplets are sprayed onto hot surfaces—has proven remarkably effective, but optimizing the process has been incredibly complex.

What makes this research particularly fascinating is its origin story. Cheng, who had played Go since high school, became intrigued when an AI-powered version called AlphaGo defeated professional human players. Rather than feeling discouraged when the AI consistently beat him, Cheng saw opportunity. He recognized that the interconnected dynamics of Go mirrored the complex network of interacting parameters in spray-cooling systems.

"Success—whether winning the game or optimizing the system—requires a holistic understanding of the network and careful management of its interactions," Cheng explained, highlighting how AI could analyze complex patterns and guide optimal strategies.

The science behind their innovation is equally captivating. When liquid droplets hit a hot surface, each tiny sphere evaporates and carries away heat, creating an army of microscopic cooling agents. As Lori notes, droplets work exceptionally well because they boil and evaporate so quickly, allowing for much more effective temperature control than traditional cooling methods.

The challenge lay in managing countless variables: optimal droplet size, nozzle hole dimensions, and which fluids work best—from water to specialized solvents and engineered mixtures. By creating AI algorithms to analyze data from 25 previous studies, the team found patterns that would have taken years to discover through traditional trial and error.

What's particularly exciting is that this research doesn't just improve one specific cooling system—it's redefining how we understand and design thermal systems for the future. The team's AI-powered approach offers valuable, ready-to-test predictions that could soon translate into real-world applications.

Looking ahead, the researchers will validate their predictions through physical testing, potentially leading to more effective ways to keep engines, computers, turbines, and other vital equipment running at optimal temperatures. In our world of increasingly powerful electronics and growing energy demands, such innovations couldn't come at a better time.

"Even though AI always wins on the Go board, I never felt frustrated but learned to take advantage of AI to tackle challenges and dilemmas in real life," Cheng reflected, embodying the optimistic, problem-solving spirit that drives scientific progress forward.