SBIR-STTR Award

This project will develop the fundamental control algorithms and perform the systems engineering studies necessary to determine the technical and economic feasibility of robotic transplanting. Animated CAD simulations of work-cells and plant materials movement are used to evaluate the capacity of commercially available robots. These computer simulations also determine the technical feasibility of each design. These results will be combined with site-specific data to determine the costs of robotic transplanting, as measured by the cost per plant\.

Applications:
A robotic transplanter could have applications in many segments of the plant production industries. Immediate applications would be transplanting bedding plants, and foliage and flowering plants. One of the final stages of tissue culture production also requires transplanting and a robot could be useful there. Since the costs of transplanting amount to hundreds of millions of dollars annually, the development of a robotic transplanter could make these segments of agriculture more profitable. Robotic materials handling, such as transplanting, could become the technology which saves the U.S. plant biotechnology industry from moving essentially all production facilities to countries with lower labor costs.

This project will develop the machine vision sensing, image processing algorithms and plant materials handling technologies necessary for high quality and economical robotic transplanting of bedding plants. Plant growth data will determine suitable ranges of characteristics for successful robotic transplanting. The direct result of this project will be a demonstrated robotic transplanter. Spinoff technologies include general purpose, low cost machine vision hardware, and machine vision and image processing algorithms for on-line, continuous monitoring of plant health\.

Applications:
This project will develop machine vision and plant materials handling technologies that provide efficient, high quality, cost effective, transplanting of bedding plants, potted plants, field crops (vegetables) and tissue cultured plants. The customized, microelectronic hardware for relevant-time machine vision has potential commercial value as a sensor for spraying weeds, guidance of directed sprays, quantifying crop residue on the soil surface, and counting, measuring, and grading agricultural products. The hardware and software developed for collecting and analysis of multispectral images of transplanted seedlings have potential as an on-line continuous monitor of plant nutritional status, including water stress.