When Jasmine first entered the wildlife reserve for research, she discovered that due to patrol blind spots, poaching incidents occurred frequently, and rare animals faced serious threats. She decided to self-study computer vision and drone control technology during the summer, independently designing and assembling an inspection drone equipped with infrared thermal imaging and deep learning models. During each test flight, she repeatedly fine-tuned detection algorithms in complex terrain and worked with rangers to improve dangerous area annotations. After dozens of iterations, the system could identify animal groups and suspicious human figures in real-time and automatically send alerts. After deployment, patrol efficiency increased by 40%, successfully preventing multiple poaching incidents, demonstrating the enormous potential of technology to protect ecosystems, and adding impressive practical experience to Jasmine's Stanford application.