Performing Department
(N/A)
Non Technical Summary
Harvesting cilantro and other similar specialty crops, such as parsley and spring onions, requires the product to be cut, bunched and tied with plastic ties showing grower information. Since there are currently no agricultural robots available that perform all of these tasks autonomously, the harvesting process is all done manually by a hand crew and it can take over 50% of the OPEX for a farm producing these specialty crops.Padma AgRobotics is developing a cilantro-harvesting robot that uses artificial intelligence (AI) and computer vision (CV) to automate the currently manual processes of: 1) cutting plants above the soil at the right height; 2) separating the plant stems; 3) forming stem bunches of adequate size and configuration; and 4) tying the bunches together for further packaging all while the machine is being driven autonomously driven forward on the plant beds. Automating the process increases productivity of the harvesting operation by a factor of four -from 20 cilantro bunches per minute to 80- while keeping the same harvesting quality standards. Automating the harvesting process reduces the cost of harvesting by as much as 50%, i.e., reducing OPEX by 25%. This also allows the harvesting crew to move from to better-paying jobs at the distribution center, thus reducing the increased costs due labor shortages. In addition, automation of the harvesting process creates a more robust supply chain to bring produce to the market.The proposed cilantro harvesting technology can be adapted to other specialty crops that are harvested in bunches: parsley, spinach, celery, chard, watercress, asparagus, and carrots. This will open the doors to U.S. farmers for new market opportunities to cultivate crops previously not considered due to their high harvesting costs. As a result, the U.S. could reduce importation rates for fresh produce and become more independent from other countries in supplying basic goods.
Animal Health Component
35%
Research Effort Categories
Basic
15%
Applied
35%
Developmental
50%
Goals / Objectives
This Phase II seeks to advance the system's technology developed in Phase 1 and validate it in relevant environments, aiming for a TRL 9 for the cilantro harvester, with a focus on optimizing the initial design and building a fully integrated cilantro harvester ready to be commercialized for farming applications. Padma AgRobotics has set the following specific objectives of the Phase II work plan:O1. Design and build ground-based cilantro accumulation and tracking system.O2. Develop semi-autonomous mobile platform.O3. Develop a label inserter and integrate the tying unit.O4. Integrate cutting and transporter unit.O5. Build multiple harvester arms and conveyor belt for transportation.Padma AgRobotics will address the following questions during this Phase II:1) Can the system cleanly accumulatethe cilantro in each row and feed to the bunching and tying unit without getting stuck in the soil due to soil height variation?2) Will the accumulation system cleanly separate out the plants in each row (in 8 rows of cilantro on a 84-inch bed) and feed to the bunching and tying system?3) Can the system cleanly collect, bunch and tie the cilantro with the manufacturer label at 1 mph on an 84-inch bed while steering itself through the bed based on the bed furrows?4) Will the platform be able to run for 8 hours daily continuously without any issues?5) Can the system cleanly collect, bunch and tie the cilantro while moving at 1 mph on a cilantro bed, for at least 200 feet, without getting stuck anywhere?6) Will the ground tracking system be able to ensure that the bunching and tying system is creating a tie consistently at 1 inch from the soil?7) Will the ground tracking system be able to consistently control the arm movement such that they do not get hit by the soil and get affected by ground height variation?8) Can the harvester system with multiple harvester arms operate on a cilantro bed at 1 mph to generate 1 cilantro bunch every 3 seconds from each row, consistently without any errors?9) Can the harvester identify the weeds in the cilantro plants and avoid bunching those sections?
Project Methods
Padma Agrobotics aims to advance the system's technology from the current prototype to a commercially viable production model. This will be achieved through an iterative design process, using field test results and feedback to make necessary adjustments and refinements, enhancing the machine's performance and reliability. The main R&D aspects to address in Phase II are:1) Ability to build a compact and robust ground tracking system2) Integrating the accumulation, tying and cutting system with the ground tracking system3) Integrating the harvesting arm into the autonomous driving platform. Once all features are integrated and tested on the test floor, the prototype will be field-tested on cilantro beds grown for the purpose to ensure it meets all requirements and specifications. Upon detecting flaws, adjustments and improvements will be iteratively made until reaching a final, reliable, optimal design. The entire iterated design process will be fully documented including the design process, specifications, test results, and any modifications made throughout the prototyping phase.Finally, this Phase II will include an assessment of the manufacturability of the final design, considering factors such as production cost, scalability, and ease of assembly.