Robotic Yard Weeder
- Atheer Alzahrani
- Anna Anderson
- Justin Jiang
- Jingyi Liang
- Brendan Lynch
- Matthew Malsam
Project Sponsor: Professor Albin Gasiewski
Dandelion weeding proves to be time-intensive, inefficient and negative on the environment due to implementing substantial amounts of herbicides with limited precision. Thus, homeowners often face the task of ridding dandelions on their lawns or have to make the difficult decision to let the dandelions grow rampantly.
In this interest, the Robotic Yard Weeder for personal lawn use helps reduce the time and effort that homeowners spend on maintaining dandelions, and gives homeowners a dandelion-free lawn without the back-breaking, manual work. Also, the precise application of the relatively benign vinegar solution significantly decreases the amount of toxic herbicides currently used in many commercial practices.
To specify, this identification system focuses on pinpointing dandelions using RF and image-processing. Our eradication system issues a small dose of strong vinegar (mere 20% solution), treating the dandelion. Finally, our traversal system should be to avoid lawn boundaries and perform object avoidance.
Previous iterations of this project and industrial products typically focus on large-scale agriculture and gardens and not personal lawn use. Some large-scale products identify all plants, taller than an inch, not as a weed. On the other hand, the Robotic Yard Weeder incorporates advanced neural network technology to precisely determine dandelion location in personal lawns with image processing.
The Robotic Yard Weeder can identify dandelions with neural nets, eradicate dandelions with vinegar solution, and traverse lawns with object, boundary avoidance. All in all, the Robotic Yard Weeder is environment friendly and efficient in dandelion removal, making entire lawns viable for automated dandelion weeding.