AI-driven autonomous spraying for precision weed management in specialty crop production

Abstract

Effective weed control in vegetable crops is crucial for optimizing yield and reducing herbicide use. This study developed a low-cost AI-driven robotic smart sprayer for targeting weeds in pepper and tomato crops on raised beds. Field tests evaluated its performance, with Tests 1 and 2 conducted on pepper and tomato beds, respectively. The sprayer achieved 89% and 88% precision, with missed targets below 5%. Recall was 86% and 84%, demonstrating effective real-time weed targeting while avoiding crops. Results highlight the sprayer’s potential for precise, efficient weed control in vegetable fields.