Professor Ken Goldberg and AUTOLAB Win IEEE CASE Best Paper Award
Bottom row: masks created by the Plant Phenotyping network as well as the estimated bounding disks, used to evaluate plant distribution and estimate daily garden state
IEOR Professor Ken Goldberg and UC Berkeley’s AUTOLAB have done it again! The prestigious IEEE Conference on Automation Science and Engineering (CASE) recently awarded its Best Paper Award to Goldberg and his team (M. Preston, R. Parikh, S. Aeron, S. Mukherjee, S. Adebola. S. Sharma, M. Theis, and W. Teitelbaum) for their impactful research, Automated Pruning of Polyculture Plants. Polyculture farming is the practice of growing multiple crop species in the same space. In their paper, Goldberg and his co-authors present the first-ever system capable of autonomously pruning plants in a polyculture setting, which allows the inherent benefits of polyculture farming to emerge free from the challenging manual labor that has hindered its widespread adoption in industrial agriculture practices.
Automated Pruning of Polyculture Plants
Abstract: Polyculture farming has environmental advantages but requires substantially more pruning than monoculture farming. We present novel hardware and algorithms for automated pruning. Using an overhead camera to collect data from a physical scale garden testbed, the autonomous system utilizes a learned Plant Phenotyping convolutional neural network and a Bounding Disk Tracking algorithm to evaluate the individual plant distribution and estimate the state of the garden each day. From this garden state, AlphaGardenSim [1] selects plants to autonomously prune. A trained neural network detects and targets specific prune points on the plant. Two custom-designed pruning tools, compatible with a FarmBot [2] gantry system, are experimentally evaluated and execute autonomous cuts through controlled algorithms. We present results for four 60-day garden cycles. Results suggest the system can autonomously achieve 0.94 normalized plant diversity with pruning shears while maintaining an average canopy coverage of 0.84 by the end of the cycles.
Authors: Mark Presten, Yahav Avigal, Mark Theis, Satvik Sharma, Rishi Parikh, Shrey Aeron, Sandeep Mukherjee, Sebastian Oehme, Simeon Adebola, Walter Teitelbaum, Varun Kamat, Ken Goldberg.