SCALE NEUTRAL HARVEST AID SYSTEM AND SENSOR TECHNOLOGIES TO IMPROVE HARVEST EFFICIENCY AND HANDLING OF FRESH MARKET HIGHBUSH BLUEBERRIES

• Sponsoring Institution: National Institute of Food and Agriculture
• Project Status: TERMINATED
• Funding Source: OTHER GRANTS
• Reporting Frequency: Annual
• Accession No.: 1004552
• Grant No.: 2014-51181-22383
• Project No.: GEOW-2014-07897
• Proposal No.: 2014-07897
• Program Code: SCRI
• Project Start Date: Sep 1, 2014
• Project End Date: Aug 31, 2019
• Grant Year: 2014

Project Director
Li, C.

Recipient Organization
UNIVERSITY OF GEORGIA
200 D.W. BROOKS DR
ATHENS,GA 30602-5016

Performing Department
COLLEGE OF ENGINEERING

Non Technical Summary
The U.S. blueberry industry accounts for almost two thirds of the world's production, constituting an important engine of economic growth in rural communities across the nation. Despite its remarkable growth in the past three decades, a shortage of labor for hand harvesting, the increasingly high labor costs, and low harvest efficiencies are becoming bottlenecks for sustainable development of the fresh market industry. Current mechanical harvesters substantially reduce harvesting costs but still result in significant yield losses, poor fruit quality, and are unaffordable for small- to medium-sized producers. Therefore, an affordable harvesting system that maintains fruit quality and reduces ground loss would be embraced by stakeholders of all farm sizes. This Standard Research and Extension Project will greatly improve harvest efficiency and fruit quality of fresh-market highbush blueberries through a systems approach and transdisciplinary research and extension effort integrating four major themes: (i) achieving high fruit quality and low yield loss by developing an affordable and efficient semi-mechanical, ergonomically optimized harvest-aid and conveyance system; (ii) aiding accelerated breeding for mechanical harvestability by developing high-throughput phenotyping systems using imaging techniques; (iii) developing the next generation berry impact recording device to better understand and improve harvest and postharvest handling systems; and iv) describing the dynamics of potential microbial contamination in the new harvest system. A cross-cutting goal is to conduct outreach, as well as economic and ergonomic studies to promote grower adoption. This project addresses all five focus areas of the SCRI and priorities established in stakeholder surveys.

Animal Health Component

0%

Research Effort Categories

Basic

10%

Applied

40%

Developmental

50%

Classification

Knowledge Area (KA)	Subject of Investigation (SOI)	Field of Science (FOS)	Percent
402	1120	2020	60%
712	1120	1060	20%
803	1120	3010	10%
903	1120	3020	10%

Knowledge Area
903 - Communication, Education, and Information Delivery; 712 - Protect Food from Contamination by Pathogenic Microorganisms, Parasites, and Naturally Occurring Toxins; 803 - Sociological and Technological Change Affecting Individuals, Families, and Communities; 402 - Engineering Systems and Equipment;

Subject Of Investigation
1120 - Blueberry;

Field Of Science
2020 - Engineering; 1060 - Biology (whole systems); 3020 - Education; 3010 - Economics;

Keywords

ergonomics

food safety

highbush blueberry

mechanical harvest aid

sensor

Goals / Objectives
The overall goal of this proposal is to develop a transdisciplinary approach to advance harvest efficiency and enhance postharvest handling of fresh-market highbush blueberries by developing scale-neutral harvest-aid systems and advanced sensor technologies. Specific objectives are to:1. Develop high-throughput phenotyping technologies to aid the selection of southern and northern highbush blueberry genotypes suitable for semi-mechanical harvest.2. Design a new semi-mechanical harvest-aid system for efficient mechanical fruit harvesting for small- and medium-size blueberry farms.3. Develop the next-generation berry impact recording sensor and use it to improve harvest and postharvest operations through a critical understanding of mechanical impacts.4. Determine microbial contamination and critical control points along the harvest and postharvest chain with the new harvest system.5. Conduct economic and ergonomic analyses of the developed technologies and determine the implications to the rural community through outreach and technology transfer.

Project Methods
Objective 1. Develop high-throughput phenotyping technologies to aid the selection of genotypes suitable for semi-mechanical harvestA line-scan hyperspectral imaging system (500-1000 nm) developed in PI Li's lab will be used to acquire scattering spectral images of blueberry fruit to determine firmness. The second instrument (the laser air-puff) will be used to develop a new texture index (in addition to firmness) to describe fruit resilience/elasticity. Three-dimensional imaging methods using 3D cameras will be developed to measure berry size (diameter and volume) and cluster tightness. The tightness of a cluster of blueberry fruit will be evaluated using the 3D camera at night to avoid IR light interference from the sun. Both RGB (2D) color images and depth images will be acquired, and the depth images will be used for segmentation of the fruit. A short wave near infrared (SWIR) spectral imaging system developed by PI Li will be used to acquire hyperspectral reflectance images for blueberry bruise detection in the region of 950 - 1650 nm. A dynamic thermal imaging system using pulsed-phase thermography will be developed to detect fruit bruising.Objective 2. Design a scale-neutral semi-mechanical harvesting system for efficient mechanical fruit harvesting on small- and medium-sized blueberry farmsDevelopment and testing of the semi-mechanical harvest aid system will be carried out in all four years on blueberry farms in Florida, Georgia, Mississippi, Michigan, and Oregon/Washington. A stationary platform with two levels from which workers will operate the shaking device will be built and field tested. A Campagnola pneumatic hook shaker (Brewt Power Systems, Merced, CA) and pneumatically powered BEI H-model shaker, originally electric-powered (BEI, South Haven, MI), and tethered to the platform frame, will be used to independently engage the base of each cane by workers on the lower level. At each location, trials will be conducted with two or more widely planted cultivars. Plants will be at least 3-year-old and lightly pruned for mechanical harvesting. Six 10-plant plots for each cultivar and harvest method treatment will be used to collect data on harvest efficiency, percentage mature and immature fruit removed, and fruit quality evaluation after harvest. Harvest will be performed 2 or 3 times each year at each location. Innovative conveyance systems. Two energy absorbing surfaces will be considered in this project. The first one will be an air-assisted deceleration surface. The second surface will consist of two layers of rubber-based belt (sheet with air voids between the two layers). A wireless sensing system will be developed to measure plant vibration achieved with the harvest aid system. Time study, using a stop watch, is typically used on existing jobs to measure productivity and set time standards. Recently, PI Freivalds developed the QuikTS app to be used on mobile devices to eliminate the need for a decimal-minute stopwatch, as typically used by industrial engineers. An independent comparison of hand, OTR harvesters, and the harvest-aid will be also be performed utilizing four physiological approaches: electromyograms (EMG), heart rate analysis, the Borg ratings of perceived exertion (RPE), and injury risk analyses. Postharvest fruit quality, microbial contamination, and postharvest decay development on fruit harvested with the harvest-aid system will be determined in years 2 and 3of the project and compared with that on fruit harvested by hand or with a conventional OTR harvester (Korvan 8000,Oxbo International, Lynden, WA).Objective 3. Develop the next-generation berry impact recording sensor and leverage it to improve harvest and postharvest operations.Smaller size and lighter weight. The goal is to reduce the size from 25.4 mm (BIRD I) to 19 mm for BIRD NXT (25% reduction), and the weight from 14 g to 5 g (64% reduction) through innovative engineering design. To achieve this goal, the accelerometer, microprocessor, battery, as well as the Printed Circuit Board (PCB) need to be re-designed. More powerful and user friendly software will be redesigned at both the microcontroller level (C programming language in MPLAB X IDE) and PC level (LabVIEW). A mobile App will be developed for BIRD NXT to enable the control of the sensor and access the data using mobile devices in the field. The tests will be conducted as part of the field studies to compare the three harvest methods (OTR, hand pick, and semi-mechanical harvest aid). Packinghouse operation evaluations will be performed initially with the BIRD sensor alone (before the harvest season), and subsequently with blueberry fruit on the packing line. A data acquisition system with accelerometers and compression sensors will be designed to measure mechanical impacts and compression forces experienced by blueberry fruit during field transportation. Objective 4. Determine microbial contamination and critical control points along the harvest and postharvest chain with the new harvest systemDifferent types of surfaces along the blueberry harvest chain and in the packinghouse will be sampled in order to determine the populations of natural microbial flora as well as spoilage and pathogenic microorganisms on berry contact surfaces. The sampled areas will include different parts of the harvester and equipment including stainless steel surface, rubber based belting, closed-cell polypropylene padding, plastic clamshells, and lugs. Samples will be taken at least three times throughout blueberry harvest season. Analysis of variance and means separation will be used to compare the significance of differences among bacterial populations recovered from various surfaces sampled in objective 4.1. Materials commonly used in blueberry harvesting equipment or as berry padding materials will be collected and examined for their ability to retain microorganisms as well as their clean-ability during routine sanitizing treatment. Specifically, coupons (1 × 2 cm2) made of stainless steel, rubber, polypropylene, and polystyrene will be exposed to a group of selected spoilage and pathogen microorganisms in Luria-Bertani no salt broth for 7 days at 28°C. Objective 5. Ergonomic and socio-economic analyses of the developed technologies and outreach and technology transfer to disseminate the outcomes of the projectTo ensure that the stakeholders benefit from this research and reduce farm injuries, ergonomics training program will be developed to specifically assist the growers in minimizing WRMSDs among the field pickers and the packers handling blueberry cartons. Since many of the workers may be Spanish speakers, the program will also be developed in Spanish. Much of this work will utilize the principles and techniques utilized in previously developed program for horticulture workers but will be modified to specifically target blueberry harvesting, both the purely manual hand picking and also the proposed semi-mechanical harvest-aid system. Because a simple cost-benefit analysis will not adequately address the complexity of adoption of the semi-mechanical harvesting system, we will develop an economic model tailored to the new equipment and the blueberry industry. This analysis will generate information that is expected to support the potential for commercial adoption of the new harvesting system. The development, testing, and eventual commercialization of the harvest-aid system will be carried out in close partnership with commercial equipment manufacturers (see support letters from Gregg Marrs, President of Blueline Manufacturing Company and Butch Rhoton, President of Brewt Power Systems). In addition, educational programs will be developed for growers, packers, marketing professionals, consumers and other stakeholders through various approaches.

Progress 09/01/14 to 08/31/19

Outputs
Target Audience:Target audiences include highbush blueberry growers in the United States, extension educators, engineers and scientists seeking to improve mechanical harvesting systems for fresh market blueberries. Changes/Problems: Nothing Reported What opportunities for training and professional development has the project provided?University of Georgia: Two PhD students were trained in engineering: Dr. Mengyun Zhang: A PhD student working on developing hyperspectral imaging technologies for blueberry internal bruising detection. She is working as an Associate Professor in ShiHeZi University in China. Dr. Yu Jiang, a PhD student is partially supported by the grant working on developing laser ranging sensor method for blueberry bush architecture characterization. He is working in Cornell as an Assistant Research Professor. Two MS students in engineering were also trained: Mr. Jesse Kuzy (MS), Mr. Rui Xu (MS). Four students were trained in Food Science: Ms. Himabindu Gazula (Ph.D. turn MS student): 1.0 EFT; Ms. Yumin Xu (MS student): 0.5 EFT; Ms. Kimberley Boyle (Undergraduate student): 0.5 EFT; Ms. Miranda Barr (Undergraduate student; After Ms. Boyle graduated from UGA): 0.5 EFT. 1 area extension agent/part-time Ph.D. student (Renee Allen), and 1 laboratory helper in 2017 under PI Chen and Scherm's supervision. Penn State University: 2 graduate students (Kiseok Sung and Eunsik Kim) are supported and actively working on the project. Washington State: One MS graduate student (Yixin Cai) and 1 undergraduate intern (Brenda Madrid) have worked with PI DeVetter with partial support from other block grants. Washington State and UC Berkeley: One PhD graduate student in economics Aerin Jung have worked under the supervision of the PI Gallardo and Zilberman from August 2015-May 2019. Dr. Jung was trained in conducting econometric analyses and dynamic optimization modeling. Michigan State University: one graduate research assistant (Patrick Abeli, 1 EFT), and one laboratory assistant (Benjamin Paskus) were trained. How have the results been disseminated to communities of interest?1) Three grower demonstrations of mechanical harvesting prototype were conducted in Oregon (2016), western Washington (2017) (~ 30 in attendance), and Prosser and Lynden, WA (2018). 2) 26 presentations were made to growers and stakeholders on harvest platform prototype, and other factors relevant to mechanical-assisted blueberry harvesting for fresh markets. 3) Nine newsletter/magazine articles and 3 Extension publications were published. 4) Two Blueberry Education Workshops were held at Hattiesburg, MS, in 2017 (~90 in attendance) and 2019 (59 in attendance). 5) Two in-service Training were held. A multi-state IST was held at the Southeast Fruit and Vegetable Conference in Savanah, GA on January 11, 2019. Agents from 7 states attended. Presentations on the results of the grant-funded research on mechanical harvesting of blueberry for fresh markets were made by faculty from six institutions. 6) Social media activities - 2 blog posts, 39 tweets, 7652 impressions, 258 engagements. 7) Education video - "Mechanical Harvesting of the Florida Blueberry Crop". A 6-minute educational video demonstrating the prototype harvester developed by the SCRI-funded research was developed. What do you plan to do during the next reporting period to accomplish the goals? Nothing Reported

Impacts
What was accomplished under these goals? ?Group 1: In 2015-2016, Mechanical harvesting of southern highbush blueberries in FL, GA, NC and northern highbush blueberries in MI, OR, and WA was performed. We evaluated a semi-mechanical harvesting system consisting of a tractor-pulled harvest-aid platform with fruit catching surfaces coupled with portable, hand-held, pneumatic shakers to detach fruit. The pneumatic shakers removed 3.5 to 15 times more fruit than by hand. In 2017 Mechanical harvesting of southern highbush blueberries in FL and northern highbush blueberries in CA, in OR, and in WA was performed. An Over-the-Row harvester (Oxbo 7420) was modified for research with two types of catch plates being evaluated, including a set of prototype "soft" catch plate (US Patent No. 9750188). The ergonomic analysis by electromyography showed that muscle strain was high in the shoulders and forearms especially when the shaker was operated at the shoulder height or raised. The pneumatic shakers had good selectivity on most dates. In 2018-2019, mechanical harvesting of northern highbush blueberries in Roseburg, OR, and in Lynden, WA, as well as in FL and CA was performed with modified Over-the-Row (OTR) harvesters (Oxbo 7440 and 8040). Modifications to these harvesters included the replacement of hard plastic catch plates with hollowed-out catch plates with their middle inserted with "soft" material. These modifications resulted in about 70% of fruit catching surfaces to be relatively soft and generated much less impact force compared to hard plastic surfaces. These studies demonstrated that improved fresh market quality in northern highbush blueberry is possible by using modifying OTR harvesters with Soft Catching Surfaces. The last phase of ergonomics study was to develop and evaluate an anti-vibration (AV) system and results showed that AV handles reduced overall discomfort from vibration. Group 2: 1) A new BIRD sensor was significantly improved by reducing the size by 17% and the weight by 50% compared to the previous prototype. Other improvements include a USB interface, new PC software, and an Android mobile app. The BIRD sensor was used to evaluate mechanical impact from 11 packing houses in the United States and 8 packing lines in Chile. The BIRD sensor was also used in the field to quantify the mechanical vibration of the shaker to the operators. 2) Near infrared (NIR) hyperspectral reflectance and transmittance imaging was developed for blueberry internal bruising detection nondestructively. In addition, a pulsed thermographic imaging system was developed to nondestructively detect bruised blueberries and achieved promising results. 3) An unmanned aerial system with a data processing pipeline was developed to determine morphological phenotypic traits of blueberry bushes in the field. Phenotypic traits were extracted from color images and a strong correlation was found between manual and image derived bush traits. 4) A LiDAR-based 3D imaging approach was developed to measure blueberry bush size and shape traits that are relevant to mechanical harvesting, including bush morphology (height, width, and volume), crown size, and shape descriptors. 5) Fundamental optical properties were measured for healthy and bruised blueberry tissues and the light propagation model of blueberries using Monte Carlo multi-layered simulation was conducted. These findings provide guidance to develop non-destructive sensing methods for blueberry internal bruising detection. 6) Deep learning-based techniques were developed to detect blueberry internal bruising after mechanical damage from hyperspectral transmittance images, to assess berry machine harvestability (e.g., fruit bunch compactness). Group 3: 1) We evaluated the hygienic conditions of 6 selected fresh blueberry packing lines in Georgia in 2015 and 2017. Results showed that sample site and sampling time had a significant influence on total aerobic, yeast and mold, and total coliform counts. 2) Berry lugs, rubber belts on color sorters, and premature berry disposing areas had significantly higher microbial counts than the other sites. Results suggest that some sites along fresh blueberry packing lines could become contaminated by microorganisms during packing. 3) Biofilm-forming ability of six two-strain mixtures of fecal coliforms, isolated from six different fresh market blueberry packing lines in Georgia was assessed on multiple coupons. Results emphasize the importance of selecting proper chemical sanitizers (with high efficacies in microbial inactivation and biofilm removal) and surface materials (with low tendency in attracting microbial attachments) for fresh market blueberry packing lines. 4) We determined the microbial quality of fresh blueberries collected from berry packinghouses. Sampling time had significant influence on the recovery of all three groups of microorganisms. This study suggests that sorting of blueberries before packing for the fresh market does not lead to significant reductions in microbial counts Group 4: 1) An empirical economic model was developed to assess the impact of fresh market prices, labor wages, and yield per acre on the net revenues for a field when using four different harvesting methodologies for fresh market blueberries. 2) We employ a theoretical framework to analyze the adoption of blueberry harvesters and use as empirical validation a stochastic dominance model. Results showed that adaptability of the machine varies across states and the harvest time window associated with the percentage of fruit being destined to the fresh market exerts an impact on the likelihood of adopting mechanical harvesters. 3) An empirical economic model to assess the impact of various factors on the net revenues of using mechanical harvesters and hand labor to harvest blueberries was built. Results suggested that prices of blueberries for the processing market would have to increase and prices of blueberries for the fresh market would have to decrease in order to foster adoption of mechanical harvesters. 4) An impact assessment survey on machine harvest of blueberries for fresh market was conducted in 2015 and 2016 in seven US states and one Canadian province. Survey respondents totaled 223 blueberry producers of various production sizes and scope. Overall, the survey showed that blueberry producers have great concerns with labor costs and availability, and are open to exploring mechanization as a way to mitigate the need for hand harvest labor. Group 5: 1) Three grower demonstrations of mechanical harvesting prototype were conducted in Oregon (2016), western Washington (2017) (~ 30 in attendance), and Prosser and Lynden, WA (2018). 2) 26 presentations were made to growers and stakeholders on harvest platform prototype, and other factors relevant to mechanical-assisted blueberry harvesting for fresh markets. 3) Nine newsletter/magazine articles and 3 Extension publications were published. 4) Two Blueberry Education Workshops were held at Hattiesburg, MS, in 2017 (~90 in attendance) and 2019 (59 in attendance). 5) Two in-service Training were held. A multi-state IST was held at the Southeast Fruit and Vegetable Conference in Savanah, GA on January 11, 2019. Agents from 7 states attended. Presentations on the results of the grant-funded research on mechanical harvesting of blueberry for fresh markets were made by faculty from six institutions. 6) Social media activities - 2 blog posts, 39 tweets, 7652 impressions, 258 engagements. 7) Education video - "Mechanical Harvesting of the Florida Blueberry Crop". A 6-minute educational video demonstrating the prototype harvester developed by the SCRI-funded research was developed.

Publications

• Type: Journal Articles Status: Published Year Published: 2019 Citation: DeVetter, L.W.*, W.Q. Yang, F. Takeda, S. Korthuis, and C. Li. 2019. Modified over-the-row machine harvesters to improve northern highbush blueberry fresh fruit quality. Agriculture doi:10.3390/agriculture9010013. Invited article.
• Type: Journal Articles Status: Published Year Published: 2018 Citation: Gallardo, R.K. and J. Sauer. 2018. Adoption of Labor-saving Technologies in Agriculture. Annual Review of Resource Economics, 10: 185-206.
• Type: Journal Articles Status: Published Year Published: 2019 Citation: Gazula, H., J. Quansah, H. Scherm, C. Li, F. Takeda, P. Wang, and J. Chen. 2019. Ease of biofilm accumulation, and efficacy of sanitizing treatments in removing the biofilms formed, on selected surface coupons. Food Control. 104:167-173. Quansah, J., H. Gazula, R. Allen, H. Scherm, C. Li, F. Takeda, and J. Chen. 2019. Microbial quality of blueberries for the fresh market. Food Control. 100:92-96. Gazula, H., J. Quansah, R. Allen, H. Scherm, C. Li, F. Takeda, and J. Chen. 2019. Microbial loads of fresh blueberry packing lines. Food Control. 100:315-320.
• Type: Journal Articles Status: Published Year Published: 2018 Citation: Kim, E., Freivalds, A., Takeda, F., & Li, C. (2018). Ergonomic evaluation of current advancements in blueberry harvesting. Agronomy, 8(11), 266.
• Type: Journal Articles Status: Published Year Published: 2018 Citation: Jiang, Y., C. Li, F. Takeda, E. Kramer, H. Ashrafi, J. Hunter. 2018. 3D point cloud data to quantitatively characterize size and shape of shrub crops. Horticulture Research. 6 (43), 2019. Zhang, M., C. Li, and F. Yang. 2018. Optical properties of blueberry flesh and skin and Monte Carlo multi-layered simulation of light interaction with fruit tissues. Postharvest Biology and Technology. 150 , 28-41. Fan, S., C. Li, W. Huang, and L. Chen. Data fusion of two hyperspectral imaging systems with complementary spectral sensing ranges for blueberry bruising detection. Sensors. 18 (12), 2018.
• Type: Journal Articles Status: Awaiting Publication Year Published: 2020 Citation: Zhang, M; Jiang, Y; Li, C; Yang, F. 2019. Fully convolutional networks for blueberry bruising and calyx segmentation using hyperspectral transmittance imaging. Biosystems Engineering. In Press.
• Type: Journal Articles Status: Under Review Year Published: 2020 Citation: Ni, X., C. Li, H. Jiang, F. Takeda. 2020. Deep learning image segmentation for blueberry fruit machine harvestability and yield trait extraction. Under review.

Progress 09/01/17 to 08/31/18

Outputs
Target Audience:Target audiences include highbush blueberry growers in the United States, extension educators, engineers and scientists seeking to improve mechanical harvesting systems for fresh market blueberries. Changes/Problems: Nothing Reported What opportunities for training and professional development has the project provided?Two PhD students were trained in engineering: Mengyun Zhang: A PhDstudent working on developing hyperspectral imaging technologies for blueberry internal bruising detection.Yu Jiang, a PhD student is partially supported by the grant working on developing laser ranging sensor method for blueberry bush architecture characterization. The project trained a MS and an undergraduate student in food science in 2018.One PhD graduate student in economics Aerin Jung have worked under the supervision of the PI since August 2015. Students is trained in conducting econometric analyses and dynamic optimization modeling. A few other graduate students were trained in plant sciences and mechanical engineering. How have the results been disseminated to communities of interest?Thefollowing outreach activities were conducted to disseminate the results of the project to our stakeholders: Grower field days - (1) Mechanical Harvesting Equipment for Fresh Market Blueberry, western Washington, August 23, 2017, (~ 30 in attendance). Grower workshops - (1) Blueberry Education Workshop, Hattiesburg, MS, January, 2017 (~90 in attendance). County agent in-service training and demonstration (1)- The prototype blueberry harvesting platform was demonstrated in a commercial blueberry field as part of a county agent IST for Extension agents from the University of Florida and the University of Georgia, Straughn Farms, Waldo, FL, April 27, 2017. Presentations at grower meetings (8). Newsletter/magazine articles relevant to the project (3) Social media - 11 tweets, 3,889 impressions, 122 engagements. Results were presented at national conferences (ASABE, ISHS) and in journal papers. What do you plan to do during the next reporting period to accomplish the goals?Proposed 2019 work plans: Reduce hard contact surfaces from ~30% in the prototype plate to less than 5% without sacrificing integrity and durability. Harvest UFL's advanced Southern Highbush blueberry selections that have high mechanical harvest potential (P. Munoz, UFL) with conventional and 2nd-generation modified Oxbo OTR harvesters. Sort machine-harvested blueberries into non-bruised and bruised categories with a hyperspectral imaging system and evaluate for extended postharvest shelf-life. Study the effect of bruise damage on blueberry postharvest physiology by measuring CO2 and C2H2 production and quality change and berry softening during cold storage. Harvest Northern Highbush blueberries in OR and WA with 2nd-generation modified Oxbo OTR harvesters. Assess food safety concerns related to mechanically harvesting blueberries with 2nd-generation modified Oxbo OTR harvesters with components selected from FDA-listed materials for use with food. Develop information on energy transmission from shaking rods to branches and clusters with a BIRD sensor and berry detachment in several cultivars. Develop information to relate the extent of bruise damage to shelf life in major fresh-pack cultivars by using the BIRD sensor and hyperspectral imaging system to create different degrees of impact force and bruise damage. Separate damaged fruit into several bruise-damage categories prior to putting blueberries into cold storage. Measure ground loss and yield from blueberry plants pruned for OTR mechanical harvester. Determine physical properties (resistance to impacts, fruit detachment force) and potential for hydrocooling of southern highbush blueberry cultivars. Continue to develop BIRD NXT sensor. Assess additional fruit catching materials for their abilities to attract microbial cells and ease to clean by commercial sanitizers Compare the microbial quality of blueberries harvested by hands, harvest aid and machine harvesters Complement the analyses utilizing a dynamic optimization approach that includes two sources of uncertainty (e.g., labor cost and output price) to estimate the optimal adoption timing.

Impacts
What was accomplished under these goals? Group 1: In 2018, mechanical harvesting of 'Duke' and 'Draper' at two farms in OR (Roseburg and Beaverton) and 'Draper' and 'Liberty' at two farms in WA (Prosser and Lynden) was performed. All harvesting was done with over-the-row (OTR) Oxbo 7440 and 8040 harvesters with two rotary drum shakers and modified fruit catching surfaces. The 2018 studies showed that as much as 92% of blueberries harvested by rotary drum shakers and collected on soft catching surface were sorted as fresh market pack-out. Generally, fresh market pack-out percentage, fruit firmness and extent of bruise damage was slightly to moderately worse than hand-harvested fruit. The modified berry catchers had ~70% of their surface with soft neoprene sheets and the remaining 30% consisted of hard, steel surfaces. Further improvements in catch plate design are necessary to reduce the hard surface area. Machine harvested blueberries were sorted on commercial packing lines. We analyzed blueberries taken at the hopper, after a color/soft sorter, and after MAF optical sorter for bruise damage. Bruise damage evaluation performed with the electronic method and visually assessing for bruise damage in sliced blueberries were highly correlated. The hyperspectral imaging system revealed the exact location and size of impact damage. In 2018, ergonomic evaluations were conducted on completely mechanized harvesting system using the Oxbo 8040 and 7440 machines and some hand harvesting. Greatest muscle loading in the arms and shoulders was recorded for both Oxbo 8040 and 7440 harvesters, however, peak values exceeded 40% MVC for short periods of time, and the loading level was acceptable for an 8-hour day. Group 2: Sensors and instrumentation. 1) We developed a laser ranging sensor based 3D imaging approach to measure blueberry bush size and shape traits that are relevant to mechanical harvesting. 3D point clouds were acquired for 367 bushes from five genotype groups. Point cloud data were processed for characterizing bush architecture, including bush morphology (height, width, and volume), crown size, and shape descriptors.The processing pipeline of 3D point cloud data developed in this study is an effective tool for blueberry breeding programs (e.g., for mechanical harvesting) and farm management. 2) We investigated the light propagation model of blueberries using Monte Carlo multi-layered simulation. The simulation results revealed that the near infrared spectral region is an effective spectral range for inspecting bruised blueberries using either reflectance or transmittance method. 3) A push broom based hyperspectral imaging (HSI) system and a liquid crystal tunable filter based HSI system with different sensing ranges and detectors were investigated to jointly detect blueberry internal bruising in the lab using three data fusion strategies. 4) We explored a deep learning technique (FCN) to detect blueberry internal bruising after mechanical damage from hyperspectral transmittance images. 5) PI Li's lab also performed non-destructive bruising detection using our developed imaging methods in North Carolina and Oregon in summer 2018. Group 3: Berry quality and safety.1) Biofilms and fecal bacterial cells accumulated on materials as potential fruit-catching surfaces of the semi-automatic harvest aid. In 2018, we evaluated the potentials of 3 different materials (spongy, rubber-like, and fabric) as fruit-catching surfaces of the semi-automatic harvest aid. We quantified the biofilm mass and fecal coliform cells accumulated on the 3 types of materials and found that the spongy material had significantly more accumulated biofilm mass and fecal coliform bacterial cells than the other two materials and the fabric material had significantly less biofilms and bacterial cells compared to the other two surfaces. But all 3 materials had more microbial buildups than stainless steel. Most recently, another material was provided to us, and more systemic evaluations are on the way. 2)Cross resistance of fecal coliforms to chemical sanitizers and antibiotics. In 2018, we also studied the cross resistance of fecal coliforms, isolated from blueberry packing lines, to the chemical sanitizers, commonly used by blueberry packer, and the antibiotics used in treating human gastrointestinal infections. Fecal coliform isolates, two from each of the 6 blueberry packing lines, were first adapted in sublethal concentrations of chlorine dioxide. The minimal inhibitory concentrations of the antibiotic against sanitizer-adapted and non-adpated fecal colifom cells were subsequently compared. Results showed that pre-adaptation with sublethal concentrations of chlorine dioxide in general did not enhance the resistance of fecal coliforms to 15 different antibiotics used in human medicine, with only 2 exception. The pre-adaption increased the minimal inhibitory concentration of kanamycin against 2 of the fecal colifom isolates from 8 to 16 µg/ml, and of nalixic acid against a fecal coliform isolate from 16 to 32 µg/ml. Group 4: Economics. 1)Static hand-held analyses across states. An empirical economic model was developed to assess the impact of fresh market prices, labor wages, and yield per acre on the net revenues for a field when using four different harvesting methodologies for fresh market blueberries: hand held shaker, over the row (OTR), modified over the row (MOTR), and manual harvesting. The analysis was applied by separating to every major in volume producing region in the United States to analyze how different net revenues according to the harvesting method, variation across regions, and therefore assess based on each state conditions where each technology would be more likely adopted. Results are being validated and tested. 2) Option value model. The static analysis in (1) was complemented with an option value model, to estimate the value in delaying adoption due to the uncertainties faced by decision makers, especially labor cost and output price. 3) Impact assessment survey. A second version of the impact assessment survey will be conducted in the 2018-2019 season. Group 5: Outreach and Extension 2017 and 2018.Information on the prototype of the modified blueberry harvester and other information on experimental approaches to improving quality of machine-harvested blueberries for fresh markets were made available to growers through a variety of methods including meetings (field days, workshops, in-service trainings, and grower seminars), publications (peer-reviewed journal, magazine and newsletter), and social media. A survey on machine harvest of blueberries for fresh market was conducted at the beginning of the grant period in 2015 and 2016 in seven US states and one Canadian province. Overall, the survey showed that blueberry producers have great concerns with labor costs and availability, and that they are open to exploring mechanization as a way of mitigating the need for hand harvest labor. Results of the survey have been published in HortTechnology. The specific outreach activities for 2017 and 2018 are described in more detail below. Future outreach plans include: 1) an additional in-service training (IST) for key county and statewide blueberry faculty throughout the U.S. This IST is scheduled for January, 2018 at the Southeast Regional Fruit and Vegetable Conference, Savannah, GA; 2) a follow-up grower survey to determine if grower attitudes and knowledge about machine harvest of blueberries for fresh markets has changed; 3) presentations on our research results at grower meetings and field days, and professional conferences; 4) additional publications (popular and scientific articles) highlighting results from our research; and 5) educational videos and social media explaining and publicizing our results.

Publications

• Type: Journal Articles Status: Published Year Published: 2018 Citation: Gallardo, K., E. Stafne, L.W. DeVetter*, Q. Zhang, C. Li, F. Takeda, J. Williamson, W. Yang, R. Beaudry, B. Cline, R. Allen. 2018. Blueberry producers attitudes toward harvest mechanization for fresh market. HortTechnology. 28:10-16.
• Type: Journal Articles Status: Published Year Published: 2017 Citation: Fumiomi Takeda, Wei Q. Yang, Changying Li, Andris Freivalds, Kiseok Sung, Rui Xu, Bo Hu, Jeffrey Williamson and Steven Sargent. (2017). Applying new technologies to transform blueberry harvesting. Agronomy 7(2):33. doi:10.3390/agronomy7020033
• Type: Journal Articles Status: Published Year Published: 2017 Citation: Patrick, A., and C. Li*, 2017. High throughput phenotyping of blueberry bush morphological traits using unmanned aerial systems. Remote Sensing. 9(12), 1250, 2017
• Type: Journal Articles Status: Published Year Published: 2017 Citation: Kuzy, J., Y. Jiang, and C. Li*, 2017. Blueberry bruise detection by pulsed thermographic imaging. Postharvest Biology and Technology, 136 (2018): 166-177.
• Type: Journal Articles Status: Published Year Published: 2017 Citation: Fan, S., C. Li*, W. Huang, and L. Chen. 2017. Detection of blueberry internal bruising over time using NIR hyperspectral reflectance imaging with optimum wavelengths. Postharvest Biology and Technology, 134:55-66.
• Type: Journal Articles Status: Under Review Year Published: 2018 Citation: Fan, S., C. Li, W. Huang, and L. Chen. Data fusion of two hyperspectral imaging systems with complementary spectral sensing ranges for blueberry bruising detection. Sensors. Under review.
• Type: Journal Articles Status: Under Review Year Published: 2018 Citation: Jiang, Y., C. Li, F. Takeda, E. Kramer, H. Ashrafi, J. Hunter. 2018. 3D point cloud data to quantitatively characterize size and shape of shrub crops. Horticulture Research. Under review.
• Type: Journal Articles Status: Under Review Year Published: 2018 Citation: Zhang, M., C. Li, and F. Yang. Optical properties of blueberry flesh and skin and Monte Carlo multi-layered simulation of light interaction with fruit tissues. Postharvest Biology and Technology. Under review after revision.
• Type: Journal Articles Status: Published Year Published: 2017 Citation: Zhang, M., C. Li, F. Takeda, and F. Yang. 2017. Detection of internally bruised blueberries using hyperspectral transmittance imaging. Transactions of ASABE, 60(5): 1-14.
• Type: Journal Articles Status: Published Year Published: 2018 Citation: Gallardo, R.K., E. Stafne, L. Wasko DeVetter, Q. Zhang, C. Li, F. Takeda, J. Williamson, W. Yang, R. Beaudry, W. Cline, and R. Allen. 2018. Blueberry Producers Attitudes toward Harvest Mechanization for Fresh Market. HortTechnology 28:10-16.
• Type: Journal Articles Status: Published Year Published: 2018 Citation: Gallardo, R.K., and J. Sauer. 2018. Adoption of Labor-Saving Technologies in Agriculture. Annual Review of Resource Economics 10:185-206.
• Type: Journal Articles Status: Under Review Year Published: 2018 Citation: Gazula, H., and J. Chen. Ease of biofilm accumulation, and efficacy of sanitizing treatments in removing the biofilms formed, on selected surface coupons. Food Control. Under Review
• Type: Journal Articles Status: Under Review Year Published: 2018 Citation: Quansah, J., H. Gazula, R. Allen, H. Scherm, C. Li, F. Takeda, and J. Chen. Microbial quality of blueberries for the fresh market. Food Control  Under Review
• Type: Journal Articles Status: Under Review Year Published: 2018 Citation: Gazula, H., J. Quansah, R. Allen, H. Scherm, C. Li, F. Takeda, and J. Chen. 2019. Hygiene conditions of fresh blueberry packing lines. Food Control  Under Review
• Type: Conference Papers and Presentations Status: Published Year Published: 2017 Citation: Jung, A., R.K. Gallardo, and D. Zilberman. 2017. Adoption of mechanical harvester aids by the blueberry industry: Results from a dynamic optimization approach. Annual Meetings of the Agricultural and Applied Economics Association. Chicago, Il. July 31-August 2, 2017.
• Type: Conference Papers and Presentations Status: Published Year Published: 2018 Citation: Zhang, M. and C. Li. Fully convolutional networks for blueberry bruising and calyx segmentation using hyperspectral transmittance imaging. ASABE Annual International Meeting Paper No: 1801489. Detroit, Michigan. July 29-August 1, 2018.
• Type: Conference Papers and Presentations Status: Published Year Published: 2018 Citation: Takeda, Fumiomi, and Changying Li. 2018. Mechanical harvesting for fresh market. USHBC Technology Symposium. Salt Lake City, Utah, Feb. 27, 2018.
• Type: Conference Papers and Presentations Status: Published Year Published: 2017 Citation: Takeda, F. Applying new technologies to transform blueberry harvesting. Florida Blueberry Growers Association Spring Short Course. Plant City, FL. March 2017.
• Type: Conference Papers and Presentations Status: Published Year Published: 2017 Citation: Takeda, F. Applying new technologies to transform blueberry harvesting and fruit quality. Produce Marketing Association Fresh Connection-Chile. Santiago, Chile. August 2017.
• Type: Conference Papers and Presentations Status: Published Year Published: 2016 Citation: Takeda, F. Applying new technologies to transform blueberry harvesting. Berry Gardens Technical Conference. Ashford International, United Kingdom. November 2016.
• Type: Conference Papers and Presentations Status: Published Year Published: 2017 Citation: Takeda, F. Mechanical harvesting of blueberries: Applying new technologies to transform blueberry harvesting and fruit quality. Oregon Blueberry Conference. Salem, OR. February 2017.
• Type: Conference Papers and Presentations Status: Published Year Published: 2017 Citation: Takeda, F. Mechanical harvesting methods for improving fresh quality of blueberries: Applying new technologies to transform blueberry harvesting and fruit quality. Florida Blueberry Growers Association Spring Short Course. Plant City, FL. October 2017.
• Type: Conference Papers and Presentations Status: Published Year Published: 2017 Citation: Gazula, H. and J. Chen. Efficacy of sanitizers in inactivating fecal coliforms in cell cultures and on coupons made from blueberry contact surface materials. GA Bio Innovation Summit. Atlanta, GA. p. 39. 24 October 2017.
• Type: Conference Papers and Presentations Status: Published Year Published: 2017 Citation: Gazula, H. and J. Chen. Efficacy of sanitizers in inactivating fecal coliforms in cell cultures and on coupons made from blueberry contact surface materials. Int. Assn. Food Prot. Annu. Mtg. Prog. Abstr. Book. Tampa, FL. p.74. 9-12 July.
• Type: Conference Papers and Presentations Status: Published Year Published: 2017 Citation: Gazula, H., J. Quansah, R. Holland, Y. Cui1, H. Scherm, C. Li, F. Takeda, and J. Chen. Hygiene conditions of fresh blueberry processing lines. Southeast Regional Fruit and Vegetable Conference Abstr. Book. Savannah GA. 5-8 January.
• Type: Conference Papers and Presentations Status: Published Year Published: 2018 Citation: Takeda, Fumiomi. Mechanical harvesting methods for improving fresh quality of blueberries. Oregon Blueberry Conference, Salem, OR Feb.19, 2018.
• Type: Conference Papers and Presentations Status: Published Year Published: 2018 Citation: Takeda, Fumiomi. Ways to reduce bruising in the packing house. Washington State University Postharvest Fruit School, Prosser, WA, March 20, 2018.
• Type: Conference Papers and Presentations Status: Published Year Published: 2018 Citation: Takeda, Fumiomi. Using the BIRD to quantify handling damage. Washington State University Postharvest Fruit School, Prosser, WA, March 20, 2018.
• Type: Conference Papers and Presentations Status: Published Year Published: 2018 Citation: DeVetter, Lisa. Will present information on 2018 field trials. Washington Small Fruit Conference. Lynden, WA. , November 27, 2018.
• Type: Conference Papers and Presentations Status: Published Year Published: 2018 Citation: Yang, Wei. Machine harvest for fresh market blueberry research update. Oregon Blueberry Commission Research Roundup. Salem, OR, January 8, 2018.
• Type: Other Status: Published Year Published: 2018 Citation: Gazula, H., and J. Chen. 2018. Ease of biofilm accumulation, and efficacy of sanitizing treatments in removing the biofilms formed, on selected abiotic surfaces. Int. Assn. Food Prot. Annu. Mtg. Prog. Abstr. Book. 8-11 July. Salt Lake City UT p.68.
• Type: Other Status: Published Year Published: 2018 Citation: Gazula, H., and J. Chen. 2018. Ease of biofilm accumulation, and efficacy of sanitizing treatments in removing the biofilms formed, on selected abiotic surfaces. ASM Microbes. 7-11 June. Atlanta GA poster no: 7387.
• Type: Other Status: Published Year Published: 2017 Citation: Stafne, E.T. 2017. Agenda for the 2017 Blueberry Education Workshop. In: E.T. Stafne (ed.). Mississippi Vaccinium J. vol. 6(1):2.
• Type: Other Status: Published Year Published: 2017 Citation: Stafne, E.T. 2017. Is it time to switch to southern highbush? In: E.T. Stafne (ed.). Mississippi Vaccinium J. vol. 6(3):2.
• Type: Other Status: Published Year Published: 2018 Citation: Courtney, Ross. 2018. A ball for berries. (Blueberry Impacting Recording Devise). Good Fruit Grower, Vol. 69, No. 13, p. 12-13.
• Type: Other Status: Published Year Published: 2018 Citation: Takeda, Fumiomi. 2018. Improving mechanical harvesting of fresh-market blueberries. VSC News, September 2018, p. 12-14.
• Type: Other Status: Published Year Published: 2018 Citation: Lies, Mark. 2018. Research in mechanical harvest for fresh showing promise. Oregon Blueberry Commission Newsletter, September 2018.
• Type: Journal Articles Status: Published Year Published: 2017 Citation: Zhang, M., C. Li*, F. Takeda, and F. Yang. 2017. Detection of internally bruised blueberries using hyperspectral transmittance imaging. Transactions of ASABE, 60(5). 60.5 (2017): 1489-1502.

Progress 09/01/16 to 08/31/17

Outputs
Target Audience:Target audience of the project includes blueberry growers and packers, county extension agents, and researchers in the food industry, government and academia. Changes/Problems: Nothing Reported What opportunities for training and professional development has the project provided?University of Georgia: The project is currently training 1.5 graduate students, 1 area extension agent/part-time Ph.D. student (Renee Allen), and 1 laboratory helper in 2017 under PI Chen and Scherm's supervision. Both graduate students have the interest to become future faculty members. The project provides an excellent opportunity for the students to gain research training and laboratory experience. Meanwhile, they are given the responsibility to supervise the laboratory helper, which is an opportunity for them to learn management and supervisory skills. In addition to laboratory research, the graduate students are encouraged to get involved in food safety professional societies. The involvement will allow them to develop professional networks and get connected with their peers. In PI Li's lab, four graduate students and two visiting scholars (Yu Jiang, Rui Xu, Jesse Kuzy, Mengyun Zhang, Shuxiang Fan, Aaron Patrick) have worked on the project in 2017. Penn State University: 2 graduate students (Kiseok Sung and Eunsik Kim) are supported and actively working on the project. One PhD graduate student in economics (Aerin Jung) have worked under the supervision of PI Gallardo and Zilberman to conduct econometric analyses and dynamic optimization modeling since August 2015. How have the results been disseminated to communities of interest?Results of the research have been presented at the annual professional meetings, such as Southeastern vegetable and small fruit consortium, International Association for Food Protection, and Georgia Life Science Summit. In addition, the following outreach activities were conducted to disseminate the results of the project to our stakeholders: Grower field days - (1) Mechanical Harvesting Equipment for Fresh Market Blueberry, western Washington, August 23, 2017, (~ 30 in attendance). Grower workshops - (1) Blueberry Education Workshop, Hattiesburg, MS, January, 2017 (~90 in attendance). County agent In-service Training Demonstrations (1)- The prototype blueberry harvesting platform was demonstrated in a commercial blueberry field as part of a county agent IST for Extension agents from the University of Florida and the University of Georgia, Straughn Farms, Waldo, FL, April 27, 2017 Presentations at grower meetings (7) a) Pruning and training blueberry plants for machine harvesting, by Bill Cline, Florida Fall Blueberry Short Course, 26 Oct 2017, Plant City, FL (~275 in attendance). b) Machine harvesting blueberries for fresh markets: Experiences and lessons learned in North Carolina, by Bill Cline, Florida Fall Blueberry Short Course, 26 Oct 2017, Plant City, FL (~275 in attendance). c) Mechanical harvesting methods for improving fresh quality of blueberries, by Fumi Takeda, Florida Fall Blueberry Short Course, 26 Oct 2017, Plant City, FL (~275 in attendance). d) Research update: Machine harvesting for fresh quality blueberries, by Wei Qiang Yang, Oregon Blueberry Commission Research Roundup, 9 Jan 2017, Salem, OR (~30 in attendance). e) Applying new technologies to transform blueberry harvesting and fruit quality, by Fumi Takeda, Produce Marketing Association Fresh Summit, Santiago, Chile, August 4, 2017. f) Applying new technologies to transform blueberry harvesting and fruit quality, by Fumi Takeda, 2017 Oregon Blueberry Conference, Feb. 17, 2017. g) Applying new technologies to transform mechanical harvesting of fresh-market quality blueberry, by Fumi Takeda, Spring Blueberry Short Course, Plant City FL, Feb. 26, 2017. (~ 300 in attendance). Social media - 11 tweets, 3,889 impressions, 122 engagements. Newsletter/magazine articles relevant to the project (3). What do you plan to do during the next reporting period to accomplish the goals?The food quality and safety group plan to evaluate the efficacy of the sanitizers in removing biofilms from coupons made of blueberry contact materials. The instrumentation group will further develop field based high throughput phenotyping system and BIRD sensor wireless communication and recharging. The economics group will finalize the static analyses and conduct dynamic analyses comparing three harvesting methods.

Impacts
What was accomplished under these goals? Group 1: In 2017 Mechanical harvesting of southern highbush blueberries in Florida (Farthing and Meadowlark) and northern highbush blueberries in CA (Duke), in OR (Duke, Draper, and Liberty), and in WA (Draper and Aurora) was performed. An Over-the-Row harvester (Oxbo 7420) was modified for research. Two types of catch plates were evaluated. One side of the harvester retained the hard plastic catch plates while the other side we installed a set of prototype "soft" catch plate (US Patent No. 9750188) was installed. Over these catch plates, intermediate fruit deceleration surfaces were installed. One side consisted of a thin neoprene sheet while the other side a surface constructed of tarp-like sheet. At all locations, hand-held, pneumatic shakers were used to detach fruit. In WA, the OTR was modified again and two rotary head shakers were installed but the modified fruit catching surfaces were retained. Blueberry impact recording device was used to record impact force for hard and soft fruit collection surfaces. A worker operating a pneumatic shaker detached as much as 20 times for fruit than workers harvesting by hand. The average bruised area and percentage of fruit with >19% bruised was less in 'Aurora' and 'Draper' than in 'Legacy' and 'Liberty'. Fruit firmness was higher in fruit harvested by hand compared to that by pneumatic shakers in some cultivars but not significantly different in other cultivars. Although some workers were able to operate the hand-held pneumatic shakers up to an hour and detach more than 800 lb of fruit, it is very unlikely that many workers can maintain high level of physical activities longer. The ergonomic analysis by electromyography showed that muscle strain was high in the shoulders and forearms especially when the shaker was operated at the shoulder height or raised. The pneumatic shakers had good selectivity on most dates. As much as 93% blue (7% green) fruit was collected in lugs. However, on initial harvest date of Duke in CA, fruit selectivity was low (~60%), but improved steadily to >90% over a 10 day harvest duration. Group 2: Near infrared (NIR) hyperspectral reflectance imaging (950-1650 nm) with reduced spectral features was investigated for blueberry internal bruising detection at four measurement times (30 min to 12 h after mechanical impact). The overall results indicated that NIR reflectance imaging can detect blueberry internal bruising as early as 30 min after mechanical impact, and band ratio images computed from two wavelengths showed great potential to detect blueberry internal bruising on the packing line. Fundamental optical properties (absorption, reduced scattering coefficient, and scattering anisotropy) were measured for healthy and bruised blueberry tissues at the spectral range of 925-1400 nm. A single integrating sphere and inverse adding doubling method were applied to calculate the optical properties. Reduced scattering coefficient and scattering anisotropy showed significant difference between healthy and bruised blueberry tissues. A pulsed thermographic imaging system was developed to nondestructively detect bruised blueberries. Relevant features from thermal images were extracted and were demonstrated to be significantly different between healthy and bruised fruit. Classification models showed accuracies of up to 88 for Farthing and Meadowlark berries. A data processing pipeline was developed to determine morphological phenotypic traits of blueberry bushes using a low-cost unmanned aerial system in the field. Color images collected by a quadcopter are processed into three-dimensional point clouds via structure from motion algorithms. Phenotypic traits that are relevant to machine harvestability such as bush height, extents, canopy area, volume, crown diameter and width, and shape are derived and referenced to ground truth. Strong correlation was found between manual and image derived bush traits. Group 3: In 2017, two additional packing lines with 300 more total sample size (four lines with n=230 in 2015) were investigated for hygiene conditions of blueberry contact surfaces. Harvester surface materials were collected for biofilm preparation and cleaning for sanitization assays. Bacteria and yeast isolates were collected from harvester surfaces to screen and identify which isolates possess a phenotype (i.e. curli fimbriae) that would make it most successful in biofilm formation assays and subsequent cleaning and sanitization studies. A total of 80 samples were collected in the summer of 2017, trend remained the same as that of 2015 data via statistical analysis. Three sites (berry lugs, rubber belts on color sorters, and pre-mature berry disposing areas) had relatively higher microbial counts. Results shows that the lunch time samples had the highest total coliforms counts whilst the morning samples had the lowest coliform counts. Unpacked and packed blueberries were collected twice at three times from the two additional packing lines. Homogenization of berries before enumeration recovered higher yeast and mold counts and total coliform counts than rinsing For total aerobic counts and coliform counts, the morning time is significantly lower than the lunch time, and for mean yeast and mold counts, three sampling points no significantly difference. Unpacked and packed berries had statistically similar microbial counts (P>0.05). A forth sanitizer (200 ppm sodium hypochlorite (NaOCl)) were evaluated for its efficacy in killing fecal coliforms isolated from fresh blueberry packinghouses in 2017.On average, NaOCl was significantly more effective compared to all the other three sanitizers in inactivating cells of fecal coliforms in cell cultures. Group 4: An empirical economic model was developed to assess the impact of fresh market prices, labor wages, and yield per acre on the net revenues for a field when using three different harvesting methodologies for fresh market blueberries: hand held shaker, over the row (OTR), and manual harvesting. Preliminary results showed that in Florida, Georgia, and North Carolina where the prices for fresh market blueberries (with the exception of Michigan) are the highest across all states considered, the manual harvesting is the highest profitable option, compared to hand held and over the row. In Washington, Oregon, and New Jersey the hand-held device is the most profitable option. The second part of the analyses will include the option that damaged fruit is diverted to the processing market to compare how results will differ from these preliminary ones. Group 5: Information on the prototype of the modified blueberry harvester and other information on experimental approaches to improving quality of machine-harvested blueberries for fresh markets was made available to growers through a variety of methods including meetings (field days, workshops, and grower seminars), publications (peer-reviewed journal, magazine and newsletter), and social media. A survey on machine harvest of blueberries for fresh market was conducted in 2015 and 2016 in seven US states and one Canadian province. Overall, the survey showed that blueberry producers have great concerns with labor costs and availability, and are open to exploring mechanization as a way to mitigate the need for hand harvest labor.

Publications

• Type: Conference Papers and Presentations Status: Published Year Published: 2017 Citation: Takeda, F. Florida Blueberry Growers Association Spring Shortcourse. March 2017. Plant City, FL Applying new technologies to transform blueberry harvesting Takeda, F. Produce Marketing Association Fresh Connection-Chile. August 2017. Santiago, Chile. Applying new technologies to transform blueberry harvesting and fruit quality Takeda, F. Berry Gardens Technical Conference. November 2016. Ashford International, United Kingdom. Applying new technologies to transform blueberry harvesting Takeda, F. Oregon Blueberry Conference. February 2017. Salem, OR. Mechanical harvesting of blueberries: Applying new technologies to transform blueberry harvesting and fruit quality Takeda, F. Florida Blueberry Growers Association Spring Shortcourse. October 2017. Plant City, FL. Mechanical harvesting methods for improving fresh quality of blueberries: Applying new technologies to transform blueberry harvesting and fruit quality Gazula, H. and J. Chen. 2017. Efficacy of sanitizers in inactivating fecal coliforms in cell cultures and on coupons made from blueberry contact surface materials. GA Bio Innovation Summit. 24 October. Atlanta, GA. p. 39. Gazula, H. and J. Chen. 2017. Efficacy of sanitizers in inactivating fecal coliforms in cell cultures and on coupons made from blueberry contact surface materials. Int. Assn. Food Prot. Annu. Mtg. Prog. Abstr. Book. 9-12 July. Tampa, FL. p.74. Gazula, H., J. Quansah, R. Holland, Y. Cui1, H. Scherm, C. Li, F. Takeda, and J. Chen. 2017. Hygiene conditions of fresh blueberry processing lines. Southeast Regional Fruit and Vegetable Conference Abstr. Book. 5-8 January, Savannah GA.
• Type: Conference Papers and Presentations Status: Published Year Published: 2017 Citation: Zhang, M., C. Li, and S. Fan. Optical properties of healthy and bruised blueberry tissues in the near-infrared spectral region. ASABE Annual International Meeting Paper No: 1700423. Spokane, Washington. July 16-July 19, 2017.
• Type: Conference Papers and Presentations Status: Published Year Published: 2017 Citation: " Jung, A., R.K. Gallardo, and D. Zilberman. 2017. Adoption of mechanical harvester aids by the blueberry industry: Results from a dynamic optimization approach. Annual Meetings of the Agricultural and Applied Economics Association. Chicago, Il. July 31-August 2, 2017.
• Type: Other Status: Published Year Published: 2017 Citation: " Jung A., R.K. Gallardo, D. Zilberman, F. Takeda, C. Li, W. Yang, S. Sargent, and J. Williamson. Static analysis of three alternative harvesting methods for fresh market blueberries.
• Type: Other Status: Published Year Published: 2017 Citation: Stafne, E.T. 2017. Agenda for the 2017 Blueberry Education Workshop. In: E.T. Stafne (ed.). Mississippi Vaccinium J. vol. 6(1):2.
• Type: Journal Articles Status: Published Year Published: 2017 Citation: Fumiomi Takeda, Wei Q. Yang, Changying Li, Andris Freivalds, Kiseok Sung, Rui Xu, Bo Hu, Jeffrey Williamson and Steven Sargent. (2017). Applying new technologies to transform blueberry harvesting. Agronomy 7(2):33. doi:10.3390/agronomy7020033
• Type: Journal Articles Status: Accepted Year Published: 2017 Citation: Gallardo, R.K., E.T. Stafne, L. DeVetter, Q. Zhang, C. Li, F. Takeda, J. G. Williamson, W.Q. Yang, R. Beaudry, W. Cline, and R. Allen. 201x. Blueberry Producers' Attitudes toward Harvest Mechanization for Fresh Market. HortTechnology (accepted).
• Type: Journal Articles Status: Accepted Year Published: 2017 Citation: Patrick, A., and C. Li, 2017. High throughput phenotyping of blueberry bush morphological traits using unmanned aerial systems. Remote Sensing.
• Type: Journal Articles Status: Published Year Published: 2017 Citation: Kuzy, J., Y. Jiang, and C. Li, 2017. Blueberry bruise detection by pulsed thermographic imaging. Postharvest Biology and Technology, 136 (2018): 166-177.
• Type: Journal Articles Status: Published Year Published: 2017 Citation: Zhang, M., C. Li, F. Takeda, and F. Yang. 2017. Detection of internally bruised blueberries using hyperspectral transmittance imaging. Transactions of ASABE, 60(5). 60.5 (2017): 1489-1502.
• Type: Journal Articles Status: Published Year Published: 2017 Citation: Fan, S., C. Li, W. Huang, and L. Chen. 2017. Detection of blueberry internal bruising over time using NIR hyperspectral reflectance imaging with optimum wavelengths. Postharvest Biology and Technology, 134:55-66.
• Type: Other Status: Published Year Published: 2017 Citation: Stafne, E.T. 2017. Is it time to switch to southern highbush? In: E.T. Stafne (ed.). Mississippi Vaccinium J. vol. 6(3):2.
• Type: Conference Papers and Presentations Status: Published Year Published: 2017 Citation: Patrick, A., C. Li. Phenotyping morphological traits of blueberry bushes using UAS. ASABE Annual International Meeting Paper No: 1701353. Spokane, Washington. July 16-July 19, 2017.
• Type: Conference Papers and Presentations Status: Published Year Published: 2017 Citation: Fan, S., C. Li, and W. Huang. Data fusion of two hyperspectral imaging systems for blueberry bruising detection. ASABE Annual International Meeting Paper No: 1701055. Spokane, Washington. July 16-July 19, 2017.

Progress 09/01/15 to 08/31/16

Outputs
Target Audience:The primary target audience for this project include highbush blueberry growers, packers, and shippers in the United States, extension educators, engineers and scientists seeking to improve mechanical harvesting systems for fresh market blueberries. Changes/Problems: Nothing Reported What opportunities for training and professional development has the project provided?USDA Kearneysville: one high school student (Jedediah Smith) was actively involved in the project. Michigan State University. MSU professor (Eric Hanson, two industry partners, (David Trinka (Naturipe) and Mike DeGrandchamp (DeGrandchamp Blueberry), one graduate research assistant (Patrick Abeli, 1 EFT), and one laboratory assistant (Benjamin Paskus). University of Georgia: four graduate students (Yu Jiang, Rui Xu, Jesse Kuzy, Mengyun Zhang) have worked on the project under the supervision of PI Li. One Ph.D. student (Himabindu Gazula), 0.5 M.S. student (Yumin Xu) and 1 half time laboratory helper (Kimberlee Boyd) are actively involved in the food safety objective. One part-time Ph.D. student (Renee Allen) and two Ph.D. students (Himabindu Gazula and Joycelyn Quansah) have been working on the project under the supervision of PI Scherm. Two PhD graduate students in economics (Yeon A Hong and Ae Rin Jung) have worked under the supervision of the PI Gallardo and Zilberman. How have the results been disseminated to communities of interest?The results have been disseminated to the community through publications, workshops, presentations, and social media. What do you plan to do during the next reporting period to accomplish the goals?For 2017 season, a self-propelled over-the-row blueberry harvester frame which will be modified with platforms for workers operating pneumatic shakers, and with soft fruit catching surfaces will be built to assist in harvesting of blueberries with hand-held shakers working with grower collaborators in FL, MI, and Oregon. One side of the harvest-aid frame will be modified to house a new fruit catch apparatus. The following tasks related to sensors and instrumentation will be pursued in 2017: 1) further develop field based high throughput phenotyping system; 2) BIRD sensor wireless communication and recharging; 3) blueberry internal bruising early detection using reflectance mode.

Impacts
What was accomplished under these goals? Group 1: Mechanical harvesting of southern highbush blueberries in Florida, Georgia, and North Carolina and northern highbush blueberries in Michigan, Oregon, and Washington was performed. In 2016, we evaluated a semi-mechanical harvesting system consisting of a tractor-pulled harvest-aid platform with fruit catching surfaces coupled with portable, hand-held, pneumatic shakers to detach fruit. Blueberry impact recording device was used to record impact force for hard and soft fruit collection surfaces. A range of air pressures and shaker rod lengths was evaluated in a replicated studies using 'Legacy', 'Draper', 'Liberty', 'Aurora', and several southern highbush blueberry cultivars. The pneumatic shakers removed 3.5 to 15 times more fruit (g/min) than by hand. The average bruised area and percentage of fruit with >19% bruised was less in 'Aurora' and 'Draper' than in 'Legacy' and 'Liberty'. Fruit firmness was higher in fruit harvested by hand compared to that by pneumatic shakers in 'Liberty', the same in 'Aurora' and 'Legacy', and lower in 'Draper'. Compared to >25% bruise in blueberries harvested by the OTR harvester, the bruise was less than 8% in fruit harvested by hand and with pneumatic shakers. The ergonomic analysis by electromyography showed that muscle strain in the back, shoulders, and forearms was low in workers operating the light weight, pneumatic shakers that were tethered to the platform with a tool balancer. Group 2: Hyperspectral transmittance imaging method was developed to nondestructively detect internal bruises. Image analyses showed that the bruised tissues of blueberries have higher transmittance intensities than healthy tissues at the spectral range of 1000-1150 nm. It was feasible to detect blueberry bruising 30 minutes after impact regardless of whether they were stored at room temperature or refrigerated, using the stem-up orientation. An Android app for the BIRD sensor was further developed based on last year's version. The mobile application enables users to more easily view and record sensor data using mobile devices. The BIRD sensor was used in the field to quantify the mechanical vibration of the shaker to the operators. Pulse-phase thermography was investigated to detect bruised and unbruised blueberries. Results indicate that pulse-phase thermography could be a viable technique for the identification of bruised blueberries with an 80% accuracy across the entire sample. A robotic platform equipped with a Light Detection And Ranging (LiDAR) sensor was explored to measure the height of blueberry bushes in the field. Bush height could be accurately measured using point cloud with an R2 of 0.91 and an RMSE of 88.73 mm for all samples. Group 3: Five northern highbush blueberry varieties were picked with a semi-mechanical harvest and fruit catch system between late July and mid-August and evaluated for fruit quality one day after harvest and after 4 to 6 weeks in cold storage . The harvested fruit was sorted as mature and immature to determine % green fruit. Mature fruit was evaluated for firmness and extent of bruise damage after 1 day at 10 C and after 4 to 6 weeks of cold storage, plus one day at room temperature. The quality of fruit harvested by the semi-mechanical harvesting method was compared to that harvested by hand and with an over-the-row harvester. After 1 day, it appeared that in most cases there was no difference in the firmness. The net seemed to have a significant impact on the quality of Legacy fruit over the pad surface in terms of bruise damage. After cold storage, all of the observed differences virtually disappeared. Compared to the hand and machine-harvested samples, the semi-mechanical harvesting method is at least comparable to hand harvesting in Elliot and Liberty fruit in maintaining firmness during storage. The efficacy of sanitizers, commonly used by blueberry packers, in killing fecal coliform bacteria and removing the biofilm mass formed by these bacteria from berry contact surface materials were evaluated. A defined circular area (1.54 cm2) on coupons (2 X 5 cm) of the materials commonly used in blueberry packing lines and containers was inoculated with 100 µl of 5 different bacterial suspensions and each suspension consisted of 2 fecal coliforms that were isolated from each of the 5 blueberry packinghouses in our 2015 survey. Inoculated surfaces were left at 10oC to dry for 6 h. The surfaces were then treated with 100 µl of each of three sanitizers (S. Q and O) for 1 min followed by sanitizer neutralization using Dey-Engley neutralizing broth. Following the treatment, inoculated areas on coupons were swabbed and bacterial counts in swab rinse water were determined. Results showed that one of the sanitizers evaluated in the study had a significantly lower efficacy in killing the fecal coliforms and removing biofilm mass. The other two sanitizers did not show significant difference in removing biofilm mass from coupon surfaces.Microbial surface samples were collected from seven mechanical blueberry harvesters from seven different growers in Alma, Georgia. These harvesters were sampled three times a day: once before harvesting in the morning, mid-day during noon break, and in the evening. In addition, fruit that had travelled through the harvester was also collected from the different harvesters at all three time points, as well as fruit from the field prior to entering the harvester for each time point. A total of 165 harvester surface swab samples and 42 fruit samples were collected. All samples were analyzed for total aerobic plate counts, yeast and mold counts, coliform counts, fecal coliform counts and enterococcus counts. Results showed that overall microbial load tended to be higher for samples from fish scales, lower walls, conveyor belts, and lugs. There was no consistency in microbial load pattern for the three time points, morning, mid-day, and evening. Enterococci were more frequently detected than coliforms and fecal coliforms. Group 4: A theoretical and empirical economic model was built to assess the impact of various factors on the net revenues of using mechanical harvesters and hand labor to harvest blueberries. Results from an empirical economic model and a series of sensitivity analyses signal that shortening the gap between prices for the fresh and processing market would increase the likelihood of adoption, especially if prices for the fresh market drop by 26% and prices for the processing market increase by 63%. Increases in labor wages of 61% would make mechanical harvesters more profitable than hand harvesting. The traditional OTR mechanical harvester in its current incarnation is not yet a proven profitable alternative for fresh market blueberries. This study improves the understanding of the factors underlying the so-called blueberry industry expansion. Group 5: One grower field demonstration of the semi-mechanical harvesting platform prototype was conducted in Oregon. 11 presentations were presented to growers and scientists (regional and international audiences) discussing preliminary data from testing the harvest platform prototype, and other factors relevant to mechanical-assisted blueberry harvesting for fresh markets. Peer reviewed articles, newsletter/magazine articles, as well as extension publications were developed. Two news releases from Penn State University and University of Georgia were distributed. 2 blog posts and 17 tweets about our project were on the social media. Grower pre-surveys were conducted in Georgia, Mississippi, North Carolina, and British Columbia (CA), at blueberry grower meetings. In total 91 complete survey responses were obtained, 47 in Georgia, 18 in British Columbia (CA), 17 in Mississippi, and 9 in North Carolina.

Publications

• Type: Journal Articles Status: Published Year Published: 2016 Citation: Gallardo, R.K. and D. Zilberman. 2016. The Economic Feasibility of Adopting Mechanical Harvesters by the Blueberry Industry. HortTechnology 26: 299-308.
• Type: Conference Papers and Presentations Status: Published Year Published: 2016 Citation: Jiang, Y., C. Li, and F. Takeda. A Method to Non-destructively Detect and Quantify Blueberry Bruising using Hyperspectral Imaging. XI International Vaccinium Symposium, Orlando, Florida. April 13, 2016.
• Type: Conference Papers and Presentations Status: Published Year Published: 2016 Citation: Zhang, M. and C. Li. Blueberry Bruise Detection using Hyperspectral Transmittance Imaging. ASABE Annual International Meeting Paper No: 162460707. Orlando, Florida. July 17-July 20, 2016.
• Type: Conference Papers and Presentations Status: Published Year Published: 2016 Citation: Gazula, H., J. Quansah, R. Holland, Y. Cui1, H. Scherm, C. Li, F. Takeda, and J. Chen. 2016. Hygiene conditions of fresh blueberry processing lines. Int. Assn. Food Prot. Annu. Mtg. Prog. Abstr. Book. 31 July  3 August. St. Louis, MO. p. 181.
• Type: Conference Papers and Presentations Status: Published Year Published: 2016 Citation: Quansah, J., H. Gazula, R. Holland, Y. Cui, H. Scherm, C. Li, F. Takeda, and J. Chen. 2016. Microbial quality of blueberries for the fresh market. Int. Assn. Food Prot. Annu. Mtg. Prog. Abstr. Book. 31 July  3 August. St. Louis, MO. p. 181.
• Type: Journal Articles Status: Published Year Published: 2016 Citation: Jiang, Y., C. Li, and F. Takeda. 2016. Nondestructive detection and quantification of blueberry bruising using near-infrared (NIR) hyperspectral reflectance imaging. Scientific Reports 6.
• Type: Conference Papers and Presentations Status: Published Year Published: 2016 Citation: Sun, S. and C. Li. Development of a field robotic phenotyping system for blueberry canopy mapping with LiDAR. ASABE Annual International Meeting Paper No: 162461174. Orlando, Florida. July 17-July 20, 2016.
• Type: Conference Papers and Presentations Status: Published Year Published: 2016 Citation: Fumiomi Takeda, 2016. Mechanical harvesting of blueberries. Oregon Blueberry Conference. Portland, OR. Jan. 28, 2016. Approximately 300 attendees.
• Type: Conference Papers and Presentations Status: Published Year Published: 2016 Citation: Changying Li. Development and Application of Berry Impact Recording Device (BIRD) to Measure Impacts Experienced by Blueberries. January 25, 2016. Oregon Blueberry Conference, Portland, Oregon.
• Type: Conference Papers and Presentations Status: Published Year Published: 2016 Citation: Fumiomi Takeda. 2016. How to reduce fruit damage during machine harvesting and on packing line. Lower Mainland Horticultural Improvement Association  Grower Short Course, Abbotsford, British Columbia, CANADA. Jan. 30, 2016. Approximately 200 attendees.
• Type: Conference Papers and Presentations Status: Published Year Published: 2015 Citation: Fumiomi Takeda. 2015. Machine harvesting for fresh market and picking machine design. Great Lakes Fruit and Vegetable Expo. Grand Rapids, MI. Dec. 11, 2015. Approximately 300 attendees.
• Type: Conference Papers and Presentations Status: Published Year Published: 2016 Citation: Fumiomi Takeda. 2016. Semi-mechanical harvesting platform for blueberries with fresh-market quality. International Society for Horticultural Science. International Vaccinium Symposium. Lake Buena Vista, FL. April 11, 2016. Approximately 300 attendees.
• Type: Conference Papers and Presentations Status: Published Year Published: 2015 Citation: Fumiomi Takeda. 2015 Berry Gardens Growers and Technical Conference, Nov. 18, 2016. Ashford, United Kingdom. 150 in attendance. Mechanical harvesting systems for blueberries.
• Type: Other Status: Published Year Published: 2016 Citation: Galinato, S., R.K. Gallardo, and Y. Hong 2015 Cost Estimates of Establishing, Producing and Packing Organic Highbush Blueberries in Eastern Washington" TB28E, September 2016.
• Type: Other Status: Published Year Published: 2016 Citation: Jung A., R.K. Gallardo, D. Zilberman. Analyzing Factors Impacting Blueberry Prices in the United States.
• Type: Other Status: Published Year Published: 2016 Citation: Stafne, E.T. 2016. 2016 Mississippi Blueberry Education Workshop. Mississippi Vaccinium Journal 5(1):2
• Type: Other Status: Published Year Published: 2016 Citation: Stafne, E.T. 2016. Help with a short survey. Mississippi Vaccinium Journal 5(1):13
• Type: Other Status: Published Year Published: 2016 Citation: Chambers, E. and P. Krewson. 2016. Engineers work on better blueberry picker. Fruit Grower News. http://fruitgrowersnews.com/news/engineers-work-better-blueberry-picker/

Progress 09/01/14 to 08/31/15

Outputs
Target Audience:The primary target audience for this project include highbush blueberry growers, packers, and shippers in the United States, extension educators, engineers and scientists seeking to improve mechanical harvesting systems for fresh market blueberries. Changes/Problems: Nothing Reported What opportunities for training and professional development has the project provided?University of Georgia: three graduate students (Yu Jiang, Rui Xu, Jesse Kuzy) and one undergraduate student (Jamel Hunter). 1 UGA area extension agent/part-time Ph.D. student (Renee Allen), and 2 additional Ph.D. students (Himabindu Gazula and Joycelyn Quansah). Another Ph.D. student (Yue Cui) and a part-time laboratory assistant (Carol Williams) also contributed to the project. Michigan State University. 2 MSU extension educators (Mark Longstroth and Carlos ), two industry partners, (David Trinka and Rolland Groenick [MBG]), one graduate research assistant (Patrick Abeli), three visiting scholars (NIhad Alsmairat [Jordan], George Guimares [Brazil], and Rosella Briano [Italy]), and one laboratory assistant (Benjamin Paskus). Two PhD graduate students in economics have worked under the supervision of the PI (Karina Gallardo). How have the results been disseminated to communities of interest?The results have been disseminated to the community through publications, workshops, presentations, and social media. What do you plan to do during the next reporting period to accomplish the goals?For 2016 season, an over-the-row frame will be either purchased ('Easy Harvester' from BBC Corporation) or will be designed and built in-house to assist in harvesting of blueberries with hand-held shakers. Collaborators in FL/GA, NC/MI, and Oregon will receive a harvest-aid unit. One side of the harvest-aid frame will be modified to house a new fruit catch apparatus.

Impacts
What was accomplished under these goals? Group 1: Harvesting of northern highbush blueberries in Linares, Chile and Aurora, Oregon and southern highbush blueberries in Waldo, Florida was performed with hand-held, pneumatically-powered shakers and by hand. In some cultivars fruit evaluation was performed on fruit machine harvested with an over-the-row unit. For 3 hand-held shakers, a range of air pressures and shaker rod lengths were evaluated in a replicated studies using 'Legacy', 'Draper', 'Liberty', 'Aurora', and several southern highbush blueberry genotypes. Portable fruit catch frames and hand-held shakers were designed and built for field testing. Group 2: A method was developed to use a near infrared hyperspectral imaging system to non-destructively phenotype blueberry internal bruising. A new index (bruise ratio) was derived from the image classification to characterize bruising. The bruise detection accuracy using the spectra extracted from the images was over 90%. The design of the BIRD sensor was significantly improved by reducing the size by 17% and the weight by 50% compared to the previous prototype BIRD I. USB interface was used to directly connect the sensor with the computer, removing the interface box used previously. LabVIEW-based PC software was designed to configure the sensor, download and process the data. A mobile app was developed for the BIRD sensor. The mechanical impacts that occur on blueberry packing lines were measured quantitatively with the BIRD sensor at 11 packing houses in the United States and 8 packing lines in Chile. The severity of bruise damage resulting from fruit being dropped was related to the impact data recorded by the BIRD sensor. Group 3: Microbial surface samples were collected from seven mechanical blueberry harvesters (four bottom loaders and three top loaders) from seven different growers at three times a day in Georgia. Fruit samples before and after going through the harvesters were also collected. Five berry processing plants with 10 berry-contact sites in each line were sampled in Georgia. A total of 230 environmental swab samples and 92 fruit samples were collected. All samples were analyzed by dilution-plating for total aerobic plate counts, yeast and mold counts, coliform counts, fecal coliform counts and enterococcus counts. Fruit samples were processed by rinsing before microbial enumeration. Fruit from the commercial production region in western Michigan were hand harvested by commercial harvesting crews and subjected to various degrees of damage by dropping fruit from various heights multiple times onto hard and padded surfaces. These berries were sliced in half longitudinally then the amount of bruising was assessed from photos taken of the sliced berries and scored. Other berries were stored for four weeks at 0 °C and nearly 100% humidity and internal fruit quality was assessed afterwards. Group 4: A theoretical and empirical economic model to assess the impact of various factors on the net revenues of using mechanical harvesters and hand labor to harvest blueberries was built. The factors potentially impacting net revenues include blueberry prices for the fresh and processing market, output productivity, harvest-induced losses in productivity, harvest-induced losses in quality (pack-out losses), labor -picker and machine driver- wages, and machine ownership costs. Group 5: A primary outreach activity during 2015 was to develop a grower pre-survey and begin surveying blueberry growers to determine their opinions, concerns, and needs regarding mechanical harvest of blueberries from fresh markets. A pre-survey as developed in 2015 and growers from Florida, Washington, and Oregon were surveyed. Grower surveys will be conducted in other major U.S. blueberry regions during 2016. In Florida, growers were surveyed at the Spring Blueberry Short Course in Plant City, Florida on February 17. Oregon and Washington growers were surveyed at a regional blueberry school held in Corvallis, Oregon on March 17. A total of 95 responses were collected (54% from Florida, 27% from Oregon and 13% from Washington). A project website was established (http://www.scri.engr.uga.edu).

Publications

• Type: Journal Articles Status: Published Year Published: 2015 Citation: " Xu, R. and C. Li. 2015. Development of the Second Generation Berry Impact Recording Device (BIRD II). Sensors 15, no. 2: 3688-3705. " Xu, R., F. Takeda, G. Krewer, and C. Li. 2015. Measure of mechanical impacts in commercial blueberry packing lines and potential damage to blueberry fruit. Postharvest Biology and Technology. DOI: 10.1016/j.postharvbio.2015.07.013.
• Type: Conference Papers and Presentations Status: Published Year Published: 2015 Citation: " Jesse Kuzy and C. Li. 2015. ASABE paper 2015. Blueberry Bruise Detection by Pulse-Phase Thermography and Neural Network. ASABE Annual International Meeting Paper No: 2191006. New Orleans, Louisiana. July 26-July 29, 2015. " Jiang, Y. and C. Li. Non-destructive detection of internal bruising of blueberries using hyperspectral imaging. ASABE Annual International Meeting Paper No: 2189892. New Orleans, Louisiana. July 26-July 29, 2015. " Xu, R. and C. Li. Blueberry bruising evaluation using second generation Berry Impact Recording Device. ASABE Annual International Meeting Paper No: 2191006. New Orleans, Louisiana. July 26-July 29, 2015. " Gallardo, R.K. and D. Zilberman. The economics of mechanical harvesters use for blueberries for the fresh market. Poster presented at the 2015 Center for Precision and Automated Agricultural Systems Open House. Prosser, WA. September 18, 2015. " Joycelyn Quansah, Himabindu Gazula, Renee Holland, Yue Cui, Harold Scherm, Changying Li, Fumi Takeda, and Jinru Chen. 2016. Microbial quality of blueberries for the fresh market. IUFoST 2016 The 18th World Congress of Food Science and Technology, August 21-25, Dublin, Ireland. Submitted in November, 2015. " Himabindu Gazula, Joycelyn Quansah, Renee Holland, Yue Cui, Harold Scherm, Changying Li, Fumi Takeda, and Jinru Chen. 2016. Hygiene Conditions of Dry Blueberry Processing Lines. IUFoST 2016 The 18th World Congress of Food Science and Technology, August 21-25, Dublin, Ireland. Submitted in November, 2015. " Stafne, E.T. 2015. 2015 Blueberry Workshop Notice Agenda. In: E.T. Stafne (ed.). Mississippi Vaccinium J. vol. 4(1):2-3. " Stafne, E.T. 2015. Presentations from 2015 Blueberry Workshop. In: E.T. Stafne (ed.). Mississippi Vacinnium J. vol. 4(2):3. " Stafne, E.T. 2015. Mechanical Harvest Survey. In: E.T. Stafne (ed.). Mississippi Vacinnium J. vol. 4(3):1
• Type: Conference Papers and Presentations Status: Published Year Published: 2015 Citation: " Current USDA blueberry research: Seasonal shading of blueberry plants; mechanical and harvest aid systems for fresh-market pack-out blueberry; blueberry packing line evaluation  Dr. Fumiomi Takeda, USDA, Spring Blueberry Short Course, February 17, 2015, Plant City, Florida. Approximately 300 attendees. " Scale-neutral Harvest Aid System and Sensor Technologies to Improve Harvest Efficiency and Handling of Fresh-market Highbush Blueberries  Dr. E.T. Stafne, Mississippi State University, Mississippi Blueberry Growers Educational Workshop, February 12, 2015. MSU Forrest County Extension, Hattiesburg, MS. 51 Attendees. " Microbial risk assessment of blueberry harvesters was discussed by Mrs. Renee Allen, area Extension faculty for commercial blueberry production. June 5, 2015. Blueberry Research and demonstration Farm, Alma, Georgia. Approximately 15 attendees including Commissioner Gary Black. " Microbial risk assessment of blueberry harvesters was discussed by Mrs. Renee Allen, area Extension faculty for commercial blueberry production. Georgia Blueberry Growers Association meeting, August 20, 2015. Waycross Georgia. 56 growers were in attendance. " New research packing, sorting and cooling facilities; new irrigation system discussions by Maggie Schaber, blueberry breeder, and John Garner, Superintendent at the 2015 blueberry field day held at the NCSU Horticultural Crops Research Station in Castle Hayne on May 24. Approximately 40 attendees. " Evaluation of cultivars for mechanical harvestability  discussions by Bill Cline, extension specialist, and Mike Mainland, professor emeritus, at the 2015 blueberry field day held at the NCSU Horticultural Crops Research Station in Castle Hayne on May 24. Approximately 40 attendees.