Peach thinning is a major expense for peach farming operations, accounting for 60% of the total cost during annual maintenance

Advent Innovations LTD Co will develop an artificial intelligence-driven peach blossom thinning software called A-Eye through this $174,021 Project Grant from the US Department of Agriculture National Institute of Food and Agriculture. The grant falls under the Small Business Innovation Research (SBIR) Program/Small Business Technology Transfer (STTR) Program (CFDA 10.212), which aims to stimulate technological innovation to build a strong national economy by supporting scientific excellence.

Specifically, A-Eye will utilize machine vision and advanced machine learning techniques to actively identify productive peach blossoms and determine bloom density and color distribution in real-time. This will guide the thinning process and be integrated with a portable device. The software seeks to improve the efficiency and accuracy of peach thinning over manual and mechanical methods, with the goal of reducing labor costs by 60% and increasing fruit size and farm profitability. Development will take place in collaboration with the University of South Carolina and Titan Farms in Ridge Spring, South Carolina, from July 1, 2023 to February 29, 2024.

The commonly used bloom thinning methods include hand, chemical, and mechanical thinning. Hand thinning and mechanical thinning are the most widely used, but these methods are labor intensive, inaccurate, and have no built-in intelligence to properly guide the thinning process. In this project, advent innovations, in collaboration withthe university of south carolina and titan farms in ridge spring, SC, isdeveloping a smart vision and AI-driven peach thinning software (a-eye) for accurate and efficient thinning of peach blossoms. By implementing artificial intelligence, the a-eye software will be capable of real time decision making for faster and more accurate thinning, which is essential during the short thinning period available. Vision processing tools and advanced machine learning techniques will be used to actively identify productive blooms by determining the bloom density and color distribution, leading to significant time and cost savings and increased fruit size and profitability of an orchard. Unlike other purely data-driven AI methodologies, the proposed a-eye software will uniquely utilize organic information, such as human knowledgebase and experiences, along with the thinning requirements practiced by peach farms. This AI-driven visual processing unit will be integrated with a portable device (tablet, laptop, etc.) to guide the thinning process in real time.