KOALA - Shrub Pruning and Plant Health Assessment Robot

KOALA is an innovative dual-armed robot designed specifically to take on repetitive task, making it easier and safer to care for our urban greenery. In today's bustling urban areas that are experiencing rapid growth, the value of green spaces is often overlooked, yet they play a vital role in enhancing our health, comfort, and the beauty of our environments. Trees and shrubs are beneficial, but keeping them well-groomed can be challenging. Pruning in particular, is a demanding and labor-intensive job that typically involves lots of manual effort and heavy machinery, all while working outdoors in sometimes tough conditions. To tackle this issue, we introduce KOALA can detect, approach, and prune overgrown plants, like boxwood shrubs, found in parks, along sidewalks, and in public gardens.

What sets KOALA apart is its ability to navigate its surroundings all on its own, using cameras and sensors to create a detailed map that helps it figure out the best way to get the job done. It utilizes the latest technologies, including LiDAR, depth cameras, and a sophisticated object detection model called YOLOv8, to accurately identify plants and spots that need trimming. Once KOALA spots areas that require attention, it deploys its robotic arms equipped with a specially designed pruning tool. It prunes the plants with great precision to ensure they are shaped beautifully, simulating the care of a skilled horticulturist.

This remarkable robot is designed to operate in tight spaces, runs on rechargeable batteries for up to four hours, and significantly reduces the need for manual labor, all while cutting down on energy consumption and material waste. As many cities are now embracing sustainable, tech-driven solutions to everyday challenges, KOALA stands out as a groundbreaking and eco-friendly approach to maintaining the lushness and charm of urban landscapes. KOALA plays a crucial role in promoting environmental care, ensuring that residents can truly enjoy their beautiful surroundings.



Note: This Project was supported by an SUTD-ZJU Thematic Research Grant, Grant Award No. SUTD-ZJU(TR) 202205.