Progress 07/01/21 to 02/28/22
Outputs
Target Audience:Target audiences: Strawberry farmers, Cal Poly Strawberry Center Researchers, and Pest Control Advisors Changes/Problems:
Nothing Reported
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?For the remainder of the project December 2021 through February 2022 we will continue data collection, complete the development of an actuated boom to addresschallenges with maintaining distance between plants and lamps(objective 1.1), add a gasoline engine to replace the batteries powering the unit - this addresses power consumption challenges (objective 2.2), incorporate map depth feedback to georeferenced field map (objective 2.2), optimize dosing by projecting dosing onto field map (objective 2.3), complete cost estimates and final report (objective 4).
Impacts
What was accomplished under these goals?
This SBIR project aims to combine ultraviolet light (UV-C) with automation to replace existing chemical and organic treatment methods with a single non-chemical alternative that controls the biggest yield reducers in the strawberry industry. Farmers using our solution will be capable of producing food with less risk, no additional labor, and with increased profitability. Broadly this project reduces human exposure to harmful chemicals, protects the planets ecosystems, and helps maintain the long-term stability of food production for our growing world population. For commercial viability UV-C must be applied both efficiently and effectively on commercial strawberry farms. From July until November 2021 one TRIC Robotics scientist, one professional engineer, and one technician collaborated to establish a system for providing ultraviolet light as a pest control across multiple rows on a commercial strawberry farm. Multirow coverage is critical for providing pest control over enough acres to offset the cost of delivering UV-C and is challenging due to the field geometry and terrain. Further, the team developed systems to control dose, manage power, and collect field data to validate the effectiveness of UV-C across multiple rows. 1 Develop multi-array treatment capability Objective goal: Increase treatment coverage by developing a physical system that allows for modular expansion of treatment arrays. 1.1 Adapt single array prototype system for multiple UV-C array coverage Activities completed: Variables associated with the treatment arrays were distilled into algebraic expressions, then compared. Single array prototype was adapted to carry 10 modularly adjustable arrays and was tested in the field. Data collected: Data was collected to determine variation between array heights. Discussion: Shifting the array mounting rails to a longer rear frame tube distributed the vehicles weight across the rear wheels which made the robot more stable and provided more traction compared to the previous frontward weighted design. Covering multiple rows decreases treatment time, turning time, and fuel consumption. Long sequential arrays are difficult to maneuver at the end of the field but reduce the impact of errors in speed and provide flexibility in the dosing because the irradiance is spread across a longer distance. 5 or 7 row systems are favorable because these geometries are familiar to farmers and naturally fit on the farm. Single row systems pass through every row and are often obstructed by obstacles in the furrows. Variation in terrain resulted in contact between arrays and plants when the robot encountered terrain. Changes in action: To control dosing across multiple rows the boom of the robot needs to automatically adjust. Longer lamps will be used to reduce complexity and costs. 1.2 Power management and usage tracking Activities completed: Power consumption was estimated for a 10-array system and additional batteries were added to accommodate. A study on power consumption vs solar generation was conducted. Other power sources were studied and a strategy developed for commercial viability. Discussion: Lamps consume a considerable power and are the biggest constraint on the coverage. 7kWh of batteries were used on a 10-array system. Solar is impractical due to power needs. Battery costs and recharging challenges are not practical - alternative methods investigated. Changes in Action: Pivot from battery to gasoline power source. 2. Develop a system for accurately controlling UV-C dose and maximizing field coverage 2.1 Adjust vehicle trajectory to center one or more treatment arrays over planting beds Activities completed: An algorithm was developed to automatically adjust GPS waypoints to improve array centering. Data collected: GPS waypoints were mapped and then plotted. During operation waypoints that deviated from center were noted and then studied to determine a strategy for improving the array centering. Discussion: Residuals from a linear fit between GPS points were removed automatically by orthogonal projection to the best fit. This removed outlier points and better aligned the arrays with the rows. 2.2 Map depth feedback to georeferenced field map & 2.3 Optimize dosing for maximum coverage No work was completed on these objectives from July to November 2021. 3. Experimentally validate treatment capabilities in open fields 3.1 Evaluate autonomous treatment compliance Activities completed: Faults of the autonomous system were tracked and evaluated. Importantly faults and/or operator notes referring to the multirow treatment were tracked. Target velocity was compared with the differentiated GPS position data (ground truth). Data collected: GPS data and discrepancies in the treatment schedule were logged. 4 treatments disrupted due to farmer equipment blocking rows and one day there was a problem with the lamp holders. Discussion: A constant offset of 0.01 m/s from a 0.07 m/s target was observed. Velocity deviated 0.0025 - 0.005 m/s in an oscillatory behavior. This was verified by measuring the time for plants to enter the front of the treatment array and exit out of the back. Times varied between 10-11s for a 32" (0.8128m) array, corresponding to a 0.074-0.081m/s treatment time. 3.2 Evaluate dosing estimates Activities completed: Target 175 J/m2 dosing. Adapting the treatment arrays for multirow coverage led to less intense irradiation because the reflectors were centered in the row middle. Furrows were tilled to reduce terrain variations and the lamps were manually lowered. Data collected: Times were measured for plants to enter and exit the treatment array. Irradiance was measured and verified in the field. Discussion: Min and max treatment distances varied from 5-12" after tilling. Initially the treatment time was set to 12.6 seconds which resulted in a low dosing between 80-108 J/m2, then treatment was slowed to increase the dosing time to 23.4 seconds and achieve 150-200 J/m2. 3.3 Collect and analyze data Sub-objective goal: Verify control of spider mites and powdery mildew. Activities completed: Data was collected and analyzed for mites and powdery mildew. Data collected: Data was collected for mites and powdery mildew. Discussion: Spider mite data was collected prior to treatments. Interestingly white fly data demonstrated effective control. Data shows good control of both mites and powdery mildew. This season there was low pressure of both mites and powdery mildew in both UV treated and control sections. Low pressure is perhaps the biggest challenge to scaling UV technology because it is difficult to verify the effectiveness. 3.4 VerifyPhylloLux results in field conditions Sub-objective goal: Phyllolux results were verified, however, without a high pressure control it was difficult to quantify the benefit of UV-C. Activities completed: Verify dosing and results in the field by placing datalogger in the field at the height of the plants. Data collected: Dosing was verified and Phyllolux results were demonstrated. Discussion: Although irradiance was in the appropriate range it is difficult to quantify the benefits. 4. Complete cost estimates and final report No work was completed on this objective from July to November 2021
Publications
|
Progress 07/01/21 to 02/28/22
Outputs
Target Audience:Target audiences: Strawberry farmers, Cal Poly Strawberry Center Researchers, and Pest Control Advisors Changes/Problems:
Nothing Reported
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?
Nothing Reported
Impacts
What was accomplished under these goals?
? This SBIR project aims to combine ultraviolet light (UV-C) with automation to replace existing chemical and organic treatment methods with a single non-chemical alternative that controls the biggest yield reducers in the strawberry industry. Farmers using our solution will be capable of producing food with less risk, no additional labor, and with increased profitability. Broadly this project reduces human exposure to harmful chemicals, protects the planets ecosystems, and helps maintain the long-term stability of food production for our growing world population. For commercial viability UV-C must be applied both efficiently and effectively on commercial strawberry farms. From July until November 2021 one TRIC Robotics scientist, one professional engineer, and one technician collaborated to establish a system for providing ultraviolet light as a pest control across multiple rows on a commercial strawberry farm. Multirow coverage is critical for providing pest control over enough acres to offset the cost of delivering UV-C and is challenging due to the field geometry and terrain. Further, the team developed systems to control dose, manage power, and collect field data to validate the effectiveness of UV-C across multiple rows. 1 Develop multi-array treatment capability Objective goal: Increase treatment coverage by developing a physical system that allows for modular expansion of treatment arrays. 1.1 Adapt single array prototype system for multiple UV-C array coverage Activities completed: Variables associated with the treatment arrays were distilled into algebraic expressions, then compared. Single array prototype was adapted to carry 10 modularly adjustable arrays and was tested in the field. Data collected: Data was collected to determine variation between array heights. Discussion: Shifting the array mounting rails to a longer rear frame tube distributed the vehicles weight across the rear wheels which made the robot more stable and provided more traction compared to the previous frontward weighted design. Covering multiple rows decreases treatment time, turning time, and fuel consumption. Long sequential arrays are difficult to maneuver at the end of the field but reduce the impact of errors in speed and provide flexibility in the dosing because the irradiance is spread across a longer distance. 5 or 7 row systems are favorable because these geometries are familiar to farmers and naturally fit on the farm. Single row systems pass through every row and are often obstructed by obstacles in the furrows. Variation in terrain resulted in contact between arrays and plants when the robot encountered terrain. Changes in action: To control dosing across multiple rows the boom of the robot needs to automatically adjust. Longer lamp, higher power lamps were used to reduce manufacturing costs of the multiarray system and the overall number of parts needed to provide pest control. 1.2 Power management and usage tracking Activities completed: Power consumption was estimated for a 10-array system and additional batteries were added to accommodate. A study on power consumption vs solar generation was conducted. Other power sources were studied and a strategy developed for commercial viability. Discussion: Lamps consume a considerable power and are the biggest constraint on the coverage. 7kWh of batteries were used on a 10-array system. Solar is impractical due to power needs. Battery costs and recharging challenges are not practical - alternative methods investigated. Changes in Action: Pivot from battery to gasoline power source. 2. Develop a system for accurately controlling UV-C dose and maximizing field coverage 2.1 Adjust vehicle trajectory to center one or more treatment arrays over planting beds Activities completed: An algorithm was developed to automatically adjust GPS waypoints to improve array centering. Data collected: GPS waypoints were mapped and then plotted. During operation waypoints that deviated from center were noted and then studied to determine a strategy for improving the array centering. Discussion: Residuals from a linear fit between GPS points were removed automatically by orthogonal projection to the best fit. This removed outlier points and better aligned the arrays with the rows. 2.2 Map depth feedback to georeferenced field map & 2.3 Optimize dosing for maximum coverage Activities completed: Coupling georeferencing with depth data from an Intel realsense a 3-dimensional grid world was constructed to represent row height relative to treatment arrays. By discretizing the global field into a grid world, each point along the robots back receives a GPS derived velocity. Using irradiance values and by computing dosing time variation in dosing is visible. Visualization of data demonstrated clear areas where velocity can be optimized for maximizing coverage. Data collected: Distances were collected in real-time throughout the field and then mapped to georeferenced coordinates. Discussion: This data clearly shows how variations in the terrain and vehicle velocity impact the dosing. Now with this foundation we are able to incorporate these capabilities for automatically adjusting the boom, improving both coverage and treatment efficacy. 3. Experimentally validate treatment capabilities in open fields 3.1 Evaluate autonomous treatment compliance Activities completed: Faults of the autonomous system were tracked and evaluated. Importantly faults and/or operator notes referring to the multirow treatment were tracked. Target velocity was compared with the differentiated GPS position data (ground truth). Data collected: GPS data and discrepancies in the treatment schedule were logged. 4 treatments disrupted due to farmer equipment blocking rows and one day there was a problem with the lamp holders. Discussion: A constant offset of 0.01 m/s from a 0.07 m/s target was observed. Velocity deviated 0.0025 - 0.005 m/s in an oscillatory behavior. This was verified by measuring the time for plants to enter the front of the treatment array and exit out of the back. Times varied between 10-11s for a 32" (0.8128m) array, corresponding to a 0.074-0.081m/s treatment time. 3.2 Evaluate dosing estimates Activities completed: Target 175 J/m2 dosing. Adapting the treatment arrays for multirow coverage led to less intense irradiation because the reflectors were centered in the row middle. Furrows were tilled to reduce terrain variations and the lamps were manually lowered. Data collected: Times were measured for plants to enter and exit the treatment array. Irradiance was measured and verified in the field. Discussion: Min and max treatment distances varied from 5-12" after tilling. Initially the treatment time was set to 12.6 seconds which resulted in a low dosing between 80-108 J/m2, then treatment was slowed to increase the dosing time to 23.4 seconds and achieve 150-200 J/m2. 3.3 Collect and analyze data Sub-objective goal: Verify control of spider mites and powdery mildew. Activities completed: Data was collected and analyzed for mites and powdery mildew. Data collected: Data was collected for mites and powdery mildew. Discussion: Spider mite data was collected prior to treatments. Interestingly white fly data demonstrated effective control. Data shows good control of both mites and powdery mildew. This season there was low pressure of both mites and powdery mildew in both UV treated and control sections. 3.4 VerifyPhylloLux results in field conditions Sub-objective goal: Phyllolux results were verified, however, without a high pressure control it was difficult to quantify the benefit of UV-C. Activities completed: Verify dosing and results in the field by placing datalogger in the field at the height of the plants. Data collected: Dosing was verified and Phyllolux results were demonstrated. Discussion: Although irradiance was in the appropriate range it is difficult to quantify the benefits.
Publications
|
|