Progress 07/01/20 to 06/30/24
Outputs
Target Audience:Crawfish industry in the US is primarily centered around rice growing activities. Presently, 18 billion pounds of rice are commercially produced annually, primarily in Arkansas (40%), Texas, Louisiana, Mississippi, Missouri, and California. The US crawfish aquaculture industry is primarily sited in Louisiana due to flat terrain, water resources, heavy clay soils and natural populations to serve as broodstock. Louisiana produces some 90% of the nation's harvests of live crawfish, but major markets are found in metropolitan areas such as Little Rock, Dallas, Houston, Austin, Mobile, Jackson, Memphis and Atlanta. Extension agents, researchers and farmers in adjoining rice-growing states and as far afield as California and North Carolina are exploring crawfish cultivation in their respective states. All the crawfish farmers in these states will be the target audience. Existing and future crawfish processing plants and their related packaging and storage units, freight carriers, and retailers will also benefited from improved crawfish production rates and process economics. Changes/Problems:Hiring of the graduate student was not possible till Spring 2021 due to Covid-19 closures. The hiring was intentionally delayed till the Spring of 2022 due to uncertainties associated with university re-opening dates and in-person contact. After the university fully opened with in-person interaction, despite national advertisements and repeated attempts in reaching potential graduate students in sister institutions, we failed to identify and hire a suitable graduate student with robotic expertise or inclination. A directional shift, along with a budget change and no-cost extension was made in 2022. The biggest change was to move the graduate student wages to support multiple undergraduate robotics students in Mechanical Engineering Department at LSU. USDA officials graciously agreed and approved the no-cost extension. We made good progress and additional prototypes were developed and tested. The directional shift also gave us access to a much bigger robot/mobile manipulator in the mechanical engineering department. With the exception of field-testing of traps for assessing the live crawfish trapping efficiency, the original tasks were completed. Live trapping was not deemed necessary as the emphasis of the current project was on robotic operation. Furthermore, as the well-proven and field-tested traps with retrofitted collars were used, the need for live testing was further diminished. Apart from the diminished need for testing, renovations and partial closure of LSU AgCenter's Aquaculture Research Station's 146 experimental ponds did not facilitate the researchers to undertake the field work as originally planned. Currently, with the original robotic arm (developed on this grant), the operator must use a robotic controller (just like a game controller joystick) to maneuver the arm operation (grabbing, moving, emptying, rebaiting, resetting). A machine-vision task that was not included in the original proposal was added to the task list. The new machine-vision feature, if developed successfully, is anticipated to recognize the traps based on a RFID tags attached to the trap collar. Once the trap is recognized, the remaining operations are anticipated to be completed without the operator's intervention. Automated operations will save significant operators' time. Furthermore, the need for highly skilled workers (that can operate a joystick with high precision) will be reduced. With the foundation laid by the current NIFA seed grant, students are anticipated to work on the crawfish harvesting project even beyond the project period. Additional grant funding will be sought to support a research associate or a post-doctoral researcher who is well versed on automation and robotics What opportunities for training and professional development has the project provided?During the initial years of the project, six undergraduate students in the Department of Biological and Agricultural Engineering were tasked to design, test and build a miniature robotic arm harvester. The group worked on designing and fabricating mini crawfish traps, rigid collars for traps, 3d printed robotic gripper claws, and bait discharging system. The students also got a chance to work on robotic arm automation hardware and coding. During the last two years, several students in the Robotics lab in Mechanical Engineering department were given an opportunity to work on the full-sized prototype. The students were involved in the design, fabrication and testing. Students were also given an opportunity to work with a full-sized robotic manipulator and learn the necessary coding. A Robotnik's mobile manipulator (RB-KAIROS) was custom programmed and was used with a Universal Robots e-series arm. During the last year of the project, students worked on using advanced machine vision features to detect a tagged trap collar without human intervention. Students worked on YOLO (You Only Look Once) algorithm for detection of real-time objects. How have the results been disseminated to communities of interest?The media news and newspaper articles generated from this research caught the attention of several crawfish farmers in and around Louisiana. The members of the LA Farm Bureau's Crawfish Advisory commitee were also made aware of the developments on the NIFA project. What do you plan to do during the next reporting period to accomplish the goals?
Nothing Reported
Impacts
What was accomplished under these goals?
The overall accomplishments are broadly divided into two sub-accomplishments. The first set of accomplishments are related to a miniature robotic arm developed by Biological Engineering Students. The second set of accomplishments are associated with automated harvesting of full-scale crawfish traps with retrofitted collars. Miniature Robotic Arm for Crawfish Harvesting A group of 6 senior design students in the BAE department designed and built a mini robotic arm. This robotic arm was capable of lifting the trap, emptying the contents of the trap (crawfish), rebaiting, and resetting the trap. Joystick controller was used for this robotic arm. This prototype caught significant attention of the media and the news was widely circulated. Several newspapers and TV channels covered the story. A few links related to this prototype are listed below: https://www.fox8live.com/video/2021/08/09/robotic-arm-crawfish-harvester/ https://www.theadvocate.com/baton_rouge/news/environment/article_19d4e75e-f70a-11eb-904c-f761cd33fd9d.html https://wgno.com/news/lsu-students-design-robotic-arm-to-trap-crawfish-could-end-the-back-breaking-work/ Full-Size Crawfish Trap Automated Havesting with Retrofitted Trap Collars After making a directional shift on the workforce (use of multiple robotics undergraduate students as opposed to one graduate student), significant progress was made on the development of a full-size robotic arm. The original cage collars were too fragile for handing with a metal robotic arm. The team decided it was best to use the current and well-proven crawfish trap designs and reinforce the cage collars for suitability to be handled by a robotic arm. This decision was taken to avoid needless building and experimental testing (for crawfish trapping efficiency) of completely new crawfish trap designs. Several crush-proof cage collars were designed and fabricated with a 3D printer. The traditional cage collars were snipped, and the best cage collar was attached to several field-scale crawfish traps. The students successfully programmed a Robotnik's mobile manipulator (RB-KAIROS) with a Universal Robots e-series arm. Several designs of the gripper claws were also designed and fabricated with a 3D printer. These grippers were custom designed to fit the Universal Robots e-series arms. The students successfully demonstrated mock harvesting of the full-sized traps. Videos were recorded to finetune the robotic manipulator. With successful harvesting, re-baiting, and re-setting, all the set tasks for the current project were deemed completed. However, due to additional time attained from no-cost extension and left over funds, the PI decided to go one step above the set original goals. The new added task involved full-autonomous operation, including trap recognition through use of a machine-vision camera, harvesting, emptying and rebaiting of the trap. The team focused on detecting and localizing crawfish traps using the YOLO (v9) algorithm, with the ultimate goal of enabling autonomous manipulation of the traps using a Kuka robotic arm. YOLO (You Only Look Once) is an algorithm designed for real-time object detection. Its primary function is to identify and locate objects within an image or video frame in a single pass through a neural network, making it highly efficient for tasks that require quick, accurate detection.The team has developed a dataset for training the neural network, which involved collecting a large number of images of crawfish traps. More than 100 images have been collected in different configurations using the Zed2i camera system. The images were carefully labelled during the annotation phase to ensure accurate detection. Once the labelling was completed, the team started training the YOLO algorithm using the dataset. Although this task still remains to be completed, the student-team will continue to work on the project even after the end of the USDA Seed Grant Project. Students will be working on analyzing the performance of the system using the initial training. The next steps will be optimizing both size and the quality of the images to improve detection and localization. This iterative process will enhance the algorithm's ability to reliably detect and localize crawfish traps in real-world conditions. The PI and Dr. Corina Barbalata (Mechanical Engineering, Robotics Assistant Professor) are actively searching for larger grant opportunities to take this project to the next level. The ultimate goal of the project is to develop a harvesting arm that can work autonomously to detect a RFID tagged trap, grab, harvest, rebait and reset the crawfish trap in a field-setting.
Publications
|
Progress 07/01/22 to 06/30/23
Outputs
Target Audience:Crawfish industry in the US is primarily centered around rice growing activities. Presently, 18 billion pounds of rice are commercially produced annually, primarily in Arkansas (40%), Texas, Louisiana, Mississippi, Missouri, and California. The US crawfish aquaculture industry is primarily sited in Louisiana due to flat terrain, water resources, heavy clay soils and natural populations to serve as broodstock. Louisiana produces some 90% of the nation's harvests of live crawfish, but major markets are found in metropolitan areas such as Little Rock, Dallas, Houston, Austin, Mobile, Jackson, Memphis and Atlanta. Extension agents, researchers and farmers in adjoining rice-growing states and as far afield as California and North Carolina are exploring crawfish cultivation in their respective states. All the crawfish farmers in these states will be the target audience. Existing and future crawfish processing plants and their related packaging and storage units, freight carriers, and retailers will also benefited from improved crawfish production rates and process economics. Changes/Problems:As stated earlier, crawfish live testing (which was originally a manual/non-robotic operation) was excluded. Live trapping was not deemed necessary as the emphasis of the current project was on the robotic operation. Furthermore, as the well-proven and field-tested traps with retrofitted collars were used, the need for live testing was further diminished. Apart from the diminished need for testing, renovations and partial closure of LSU AgCenter's Aquaculture Research Station's 146 experimental ponds did not facilitate the researchers to undertake the field work within the university campus. A machine-vision task that was not included in the original proposal was added to the tasklist. The new machine-vision feature, if developed successfully, is anticipated to recognize the traps based on a RFID tags attached to the trap collar. Once the trap is recognized, the remaining operations are anticipated to be completed without the operator's intervention. Automated operations will save significant operators' time. Furthermore, the need for highly skilled workers (that can operate a joy stick with high precision) will be reduced. What opportunities for training and professional development has the project provided?Several students in Mechanical Engineering department were given an opportunity to work on protototype design, fabrication and testing. Students were also given an opportunity to work with a full-sized robotic manipulator and learn the necessary coding. How have the results been disseminated to communities of interest?
Nothing Reported
What do you plan to do during the next reporting period to accomplish the goals?With the exception of field-testing of traps for assessing the live crawfish trapping efficiency, most of the original tasks were completed. Live trapping was not deemed necessary as the emphasis of the current project was on robotic operation. Furthermore, as the well-proven and field-tested traps with retrofitted collars were used, the need for live testing was further diminished. For the next project period, emphasis is set on a newly added task of machine-vision based harvesting. The machine-vision task was not included in the original proposal. Currently, with the existing system, the operator has to use robotic controller (just like a game controller joystick) to maneuver the arm operation (grabbing, moving, emptying, rebaiting, resetting). The new machine-vision will recognize the trap based on an RFID tag attached to the trap collar. Once the trap is recognized, the remaining operations will be completed without the operator's intervention. The task of successful identification and harvesting is anticipated to take 2 semesters or more. With the foundation laid by the current NIFA seed grant, students are anticipated to work on the crawfish harvesting project even beyond the project period. Additional grant funding will be sought to support a research associate or a post-doctoral researcher who is well versed on automation and robotics.
Impacts
What was accomplished under these goals?
The 2022 progress report (period ending 6/30/2022) indicated the progress made by a group of 6 senior design students in the BAE department. The senior design group designed and built a mini robotic arm. This prototype caught the attention of the media. Several newspapers and TV channels covered the story. Some of the links are listed below: https://www.fox8live.com/video/2021/08/09/robotic-arm-crawfish-harvester/ https://www.theadvocate.com/baton_rouge/news/environment/article_19d4e75e-f70a-11eb-904c-f761cd33fd9d.html https://wgno.com/news/lsu-students-design-robotic-arm-to-trap-crawfish-could-end-the-back-breaking-work/ For the current project period (ending 6/30/2023), the emphasis was set on employing full-sized crawfish traps. Multiple undergraduate students working in the Mechanical Engineering Robotics lab were hired to undertake the tasks. The team decided it was best to use the current and well-proven crawfish trap designs and reinforce the cage collars for suitability to be handled by a robotic arm. This decision was taken to avoid needless building and experimental testing (for crawfish trapping efficiency) of completely new crawfish trap designs. Several crush-proof cage collars were designed and fabricated with a 3D printer. The traditional cage collars were snipped, and the best cage collar was attached to several field-scale crawfish traps. The students continued to work on the Robotnik's mobile manipulator (RB-KAIROS) with a Universal Robots e-series arm. Several designs of the gripper claws were also designed and fabricated with a 3D printer. These grippers were custom designed to fit the Universal Robots e-series arms. The students continued mock harvest testing in the mechanical engineering department's robotics laboratory. Videos were recorded to finetune and the trap harvesting, re-baiting, and trap re-setting operations.
Publications
|
Progress 07/01/21 to 06/30/22
Outputs
Target Audience:Crawfish industry in the US is primarily centered around rice growing activities. Presently, 18 billion pounds of rice are commercially produced annually, primarily in Arkansas (40%), Texas, Louisiana, Mississippi, Missouri, and California. The US crawfish aquaculture industry is primarily sited in Louisiana due to flat terrain, water resources, heavy clay soils and natural populations to serve as broodstock. Louisiana produces some 90% of the nation's harvests of live crawfish, but major markets are found in metropolitan areas such as Little Rock, Dallas, Houston, Austin, Mobile, Jackson, Memphis and Atlanta. Extension agents, researchers and farmers in adjoining rice-growing states and as far afield as California and North Carolina are exploring crawfish cultivation in their respective states. All the crawfish farmers in these states will be the target audience. Existing and future crawfish processing plants and their related packaging and storage units, freight carriers, and retailers will also benefited from improved crawfish production rates and process economics. Changes/Problems:Hiring of the graduate student was not possible till Spring 2021 due to Covid-19 closures. The hiring was intentionally delayed till the Spring of 2022 due to uncertainities associated with university re-operning dates and in-person contact. After the university fully opened with in-person interaction, despite national advertisements and repeated attempts in reaching potential graduate students in sister institutions, we failed to identify and hire a suitable graduate student with robotic expertise or inclination. A directional shift, along with a budget change and no-cost extension was made in 2022. The biggest change was to move the graduate student wages to support multiple undergraduate robotics students in Mechanical Engineering Department at LSU. USDA officials graciously agreed and approved my request last year. We made good progress and additional prototypes were developed and tested. The directional shift also gave us access to a much bigger robot in the mechanical engineering department. We are currently working on a task that is beyond what was initially proposed. We are making the robot detect the crawfish traps by cameras/machine vision and complete the harvesting sequence without the operator's continuous input. With busy undergraduate student schedules, the progress has been slower than I expected. The undergrads also took up summer internships in 2022. A second request was made for an additional one-year of no-cost extension. No budget modifications were requested this time. What opportunities for training and professional development has the project provided?
Nothing Reported
How have the results been disseminated to communities of interest?
Nothing Reported
What do you plan to do during the next reporting period to accomplish the goals?As most of the original tasks are completed, all emphasis is set on the newly added task of machine-vision based harvesting. Currently, with the existing system, the operator has to use robotic controller (just like a game controller joystick) to maneuver the arm operation (grabbing, moving, emptying, rebaiting, resetting). The new machine-vision will recognize the trap based on an RFID tags attached to the trap collars. The task of successful identification and harvesting is anticipated to take 2 semesters. Although there is indefinite room for refinement of the machine-vision based harvesting, significant progress is anticipated by May of 2024.
Impacts
What was accomplished under these goals?
The long-term goal of the proposed seed project is to lower production costs and improve the competitive position of the US crawfish industry. Specific objectives of the current seed project are: • Design, build, and test several new crawfish traps that: 1) are sturdier, 2) have higher trapping efficiency (and less escape) and 3) are more conducive to robotic handling • Design and built a prototype robotic arm that can lift selected trap(s) (finalists from objective one) with live crawfish, harvest the trapped crawfish, re-bait the traps and reset them in the correct vertical position. The last progress report indicated the progress made by a group of 6 senior design students in the BAE department. The senior design group designed and built a mini robotic arm. This prototype caught the attention of the media. Several newspapers and TV channels covered the story. Some of the links are listed below: https://www.fox8live.com/video/2021/08/09/robotic-arm-crawfish-harvester/ https://www.theadvocate.com/baton_rouge/news/environment/article_19d4e75e-f70a-11eb-904c-f761cd33fd9d.html https://wgno.com/news/lsu-students-design-robotic-arm-to-trap-crawfish-could-end-the-back-breaking-work/ After making a directional shift on the workforce (use of multiple robotics undergraduate students as opposed to one graduate student), significant progress was made on the development of a full-size robotic arm. The original cage collars were too fragile for handing with a metal robotic arm. New, crush-proof, cage collars were 3D printed. A Robotnik's mobile manipulator (RB-KAIROS) was custom programmed and was used with a Universal Robots e-series arm. This manipulator had the necessary torque for handling a full-size crawfish trap with retrofitted collars. A new gripper claw was also designed and 3D printed for grabbing the plastic trap collars. Several successful mock harvests were made in the mechanical engineering department's robotics laboratory. With successful harvesting, re-baiting, and re-setting (videos available for review), all the set tasks for the current project were fully completed. However, due to left over funds, I (PI) decided to go one step above the set original goals. The new added task involves full-autonomous operation, including trap recognition through use of a machine-vision camera, harvesting, emptying and rebaiting of the trap. This task is anticipated to be completed in the next 9 to 12 months.
Publications
|
Progress 07/01/20 to 06/30/21
Outputs
Target Audience:Crawfish industry in the US is primarily centered around rice growing activities. Presently, 18 billion pounds of rice are commercially produced annually, primarily in Arkansas (40%), Texas, Louisiana, Mississippi, Missouri, and California. The US crawfish aquaculture industry is primarily sited in Louisiana due to flat terrain, water resources, heavy clay soils and natural populations to serve as broodstock. Louisiana produces some 90% of the nation's harvests of live crawfish, but major markets are found in metropolitan areas such as Little Rock, Dallas, Houston, Austin, Mobile, Jackson, Memphis and Atlanta. Extension agents, researchers and farmers in adjoining rice-growing states and as far afield as California and North Carolina are exploring crawfish cultivation in their respective states. All the crawfish farmers in these states will be the target audience. Existing and future crawfish processing plants and their related packaging and storage units, freight carriers, and retailers will also benefited from improved crawfish production rates and process economics. Changes/Problems:Hiring of the graduate student was not possible till Spring 2021 due to Covid-19 closures. The hiring wasintentionally delayed during the Summer of 2021 due to uncertainities associated with university re-operning date. As the university started with in-person contact in the Fall of 2021, the search for hiring a suitable graduate student was initiated. A national advertisement was made to find a sutiable graduate student for January 2022. Due to unforseen Covid-19 disruptions, the PI plans to request no-cost extension on the project. What opportunities for training and professional development has the project provided?A total of 6 undergraduate students in the department of Biological and Agricultural Engineering Department got a chance to hone their design and engineering skills. The group worked on designing and frabricating mini crawfish traps, rigid collars for traps, 3d printedrobotic gripper claws, and bait discharging system. The students also got a chance to work on robotic arm automation hardware and coding. How have the results been disseminated to communities of interest?The media news and newspaper articles caught the attention of several crawfish farmers in and around Louisiana. The members of the LA Farm Bureau's Crawfish Advisory commitee werealso made aware of the developments on the NIFA project. What do you plan to do during the next reporting period to accomplish the goals?Based on the feedback from the minature prototype, the PI wishes to build a larger robotic arm in the next reporting period. New crawfish traps with collars that resist the crushing of the roboticarm will be designed and field tested in the next project period.
Impacts
What was accomplished under these goals?
Due to Covid-19 closures in 2020, the crawfish harvesting arm protype fabrication was not attemtped at thefull-scale. A suitable graduate student could not be identified and admitted to the Biological and Agricultural Engineering (BAE)Department for Fall 2020 (due to Covid-19 disruptions). However,the PI decided to assign the task of building a miniature version of the harvesting arm to a group of 6 senior design students in the BAE department. In the 2-semester time period (September 2020 - April 2021), a mini robotic arm was desinged and built by the senior design team. This prototype caught theattention of the media in the Summer of 2021. Several newspapers and TV channels covered the story. Some of the links are listed under "Other Products"
Publications
|
|