Performing Department
(N/A)
Non Technical Summary
Unusable crops are considered waste to be removed at the farmer's expense. Reintroduction of insect decomposers into agricultural systems can alleviate waste removal costs, rapidly generate usable products (fertilizer), and lead to new revenue streams (insect biomass). The Black Soldier Fly (BSF) is an effective decomposer of any type of organic matter, and is used worldwide for waste management, production of fertilizer (frass), and as feed for livestock, aquaculture, and pets. However, BSF rearing by the standard "batch" method (carefully measured cohorts of larvae with set quantities of waste) currently requires substantial human labor, which hinders on- farm use. An alternative "steady-state" rearing system involves continuous rearing of BSF in bioreactors and is less labor intensive, but still requires some human maintenance. These maintenance tasks are targets for introducing cost-conscious sensing and automation. The goal of this seed-grant project is to engineer, build, and test a prototype automated steady state system suitable for on-farm use. We will accomplish this goal through a productive and established collaboration among two entomologists, the inventor of the steady-state system, and an electrical/computer engineer. By automating on-farm BSF rearing, we will expand utilization of waste in agricultural systems and engineer new products that utilize materials from agriculture. Our project addresses the AFRI long-term goals to ensure efficient use of on-farm resources and increase integration of natural biological cycles and controls in agricultural systems.
Animal Health Component
0%
Research Effort Categories
Basic
20%
Applied
20%
Developmental
60%
Goals / Objectives
Insect decomposers are a missing link in our food and agricultural systems. Reintroduction of insects can help convert a linear food system, which creates unusable waste, to a "circular food system" that recycles waste to new products. The Black Soldier Fly (BSF), Hermetia illucens, is an ideal decomposer for circular food systems. BSF larvae (immatures) can turn any type of low-grade waste into body mass quickly. BSF biomass can then be harvested to use as feed for fowl or aquaculture, and to isolate high-end protein and fat, oil for biofuel and other valuable byproducts, such as chitin/chitosan and melanin. At the same time, these insects reduce the volume of waste and transform it into a nutrient-rich, compost-like substance called frass. Because of its efficacy in valorizing waste to usable products, BSF is now widely used around the world for waste management and decontamination. But until now, commercialized BSF operations have only been feasible at industrial scales, even though waste recycling and added revenue streams are needed across many sectors.Agriculture is one such sector where BSF have great potential to turn unusable outputs (e.g., unmarketable crops and some crop residues) into products that are usable on site (insect frass as fertilizer) and products that bring additional income (larvae or pupae). The frass produced by BSF larval feeding on waste is particularly useful to generate on-site. Frass contains both digested waste and insect exoskeletons made of chitin. It is rich in nitrogen and is a suitable replacement for environmentally damaging synthetic fertilizers. As it is broken down by microbial activity, the chitin in frass stimulates plant immunity against pests and pathogens and increases the abundance and diversity of beneficial microbes in the rhizosphere. This product is so versatile as a soil amendment, it is now being produced and marketed by the BSF industry.Despite the clear benefits of on-farm BSF rearing for cost savings on agricultural waste removal, fertilizer, soil and plant health maintenance, and generating additional income, BSF rearing remains confined to large, for-profit entities focused exclusively on insect rearing at industrial scales. This sector can absorb the labor costs of BSF production through the current standard "batch" method, whereby carefully measured cohorts of larvae are established with set quantities of food waste as a substrate, then maintained through development for harvest. However, BSF rearing on-farm using the batch methods adopted at industrial scales would involve manual labor inputs that would outweigh any cost savings gained through production of fertilizer or marketable larvae.An alternative that is more flexible, less labor intensive, and adaptable to on-farm scales is the "steady state" rearing method, which involves continuous rearing of BSF in bioreactors housing overlapping generations of BSF larvae. The bioreactor environment is partially open, allowing natural oviposition by adults and eliminating the need for labor intensive egg collection and weighing. The design also takes advantage of the natural "self-harvesting" behavior of mature larvae, which are known as prepupae when they stop feeding and darken. Insects in this stage remove themselves from the substrate and migrate to a waste-free and dry location (e.g., a bin) for final maturation to adults. Although less labor intensive than batch rearing, steady state rearing operations still require some human labor for maintenance. This includes dispensing of waste to bioreactors, stacking and de-stacking of bioreactor bins, removal of spent waste (frass product) to spent waste reservoirs, vacuum removal of self-harvested prepupae to collection vessels, and monitoring bioreactors for issues that would prompt corrective action (e.g., checking for excessive heat, monitoring carbon dioxide, ammonia, pH, moisture).The long-term goal of the proposed project is to automate steady-state rearing maintenance tasks that currently require human labor to enable on-farm use of BSF for agricultural waste recycling. The proposed project will address this goal through the following three objectives:Objective 1: engineer cost-conscious solutions to automate maintenance and monitoring tasksObjective 2: compare costs and returns of the automated system to more labor-intensive methodsObjective 3: evaluate and demonstrate the performance of an automated system in an agricultural context.Our efforts will produce the first prototype of a semi-automated steady-state BSF rearing operation suitable for use on small to mid-sized farms, as well as use in other contexts where food/green waste is abundant, but labor costs high (e.g., college campuses).
Project Methods
Engineer cost-conscious solutions to automate maintenance and monitoring tasks The goal of Objective 1 is to identify key processes in steady state larval farming operations that can be the target for automation and develop cost-conscious solutions toward automating those processes. All developed components will be integrated into a semi- automated prototype that will be evaluated against manual baselines. Overall, there are two families of processes to be automated, which broadly relate to maintenance and monitoring. In the maintenance category we propose developing and testing 1) a soft robotic arm that can reach in all areas of a BR2 bin without damaging the larvae, 2) a system for moving spent waste into a spent waste reservoir, aerating the latter, and discharging the processed waste from the BSF facility, and 3) replenishing each BR2 bin by metering out feedstock and bulking agent, and mixing of larvae-seeded food waste in trays during each loading cycle. We anticipate devising an integrated actuation-perception means that uses multi-modal perception (e.g., thermal and visible-spectrum imaging) to complement a feedforward replenishment of larvae- seeded food waste and bulking agents. Other maintenance tasks (like stacking and unstacking of bins integrated with the loading cycle, sweeping of egg clutches left by mating adults upon de-stacking each column during the loading/seeding cycles of the BR2 bins, clearing larval gutter channels, etc.) will take place manually at the proposed stage of development as they are processes that would require significant engineering effort that exceeds the scope of this proposed project. In the monitoring category, we seek to develop a portable multi-sensor system that measures key properties in the environment that promote growth. We will monitor pH to ensure that the pH remains neutral to alkaline where larvae grow best. We also will develop larvae activity monitoring algorithms using data from thermal cameras for mismanagement and corrective actions (larval stress indicators, such as aggregation in corners or fountain-like movement to the middle of the bins) and for estimating the number of larvae. In addition, we will monitor abiotic conditions by sensors for real-time corrective actions (humidity, temperature, CO2, ammonia volatilization, etc.). Corrective actions (add/remove liquid, operate a fan, etc.) will be achieved via a mechatronics system to be developed under this objective. Compare costs and returns of the automated system to more labor-intensive methodsThe semi-automated prototype obtained with Objective 1 activities will be compared with a fully manual steady state operation, using relevant Ag biowaste, which will be collected, homogenized, and stored during the months prior to prototype completion. Substrate inputs will include brewery's spent grains, tomato or grape pomace, and citrus fruit in set proportions. The feedstock will be accumulated over the first 6-12 months and preserved through fermentation in 55-gal drums. Fermentation will result in a stable product, allowing us to use the feedstock up to a year from collection. Each rearing operation (steady state manual or steady state automated) will consist of a BR1 and a stack of two BR2s. This will require approximately 8 kg/day of feedstock per treatment, with the feedstock being added to the bioreactors twice a week. During each iteration of the comparison, we will measure the following response variables: yield of prepupae (expressed as wet weight of prepupae/wet weight of feedstock), adult emergence rate (measured as % of empty pupal exuviae, taken on five random samples of 200 pupae per BR1), and frass yield (as wet weight of frass per weight of feedstock). We will also track labor input, expressed as hours (FTE) per unit output of products (pupae and spent waste). Labor includes all operations involved in the maintenance of the steady-state system: dispensing feedstock, clean up, maintenance, collection of prepupae, and disposal of spent waste. The experiment will run for three overlapping generations (~3 months) and will be replicated three times over year 2.Evaluate and demonstrate the performance of an automated system in an agricultural context.We will adjust the system as needed based on Objective 2, then perform an evaluation of system performance with freshly collected agricultural waste as an emulation of typical on-farm use. This last evaluation will primarily serve as a demonstration trial for relevant stakeholders (communication efforts). We will integrate a tour of the prototype with an educational lecture on BSF farming.Efforts to cause a change in knowledge, behavior or conditionsData sharing with collaborators and the general publicPresentations at academic meetingsPresentations to public and private stakeholdersMentoring of graduate and undergraduate students and postdoctoral scholarsPublication of manuscripts in open access journals EvaluationObjective 1 milestonesRealize engineering designs and physical mechatronics and robotics prototypesDevelop algorithms both for sensing and decision making for actions to be performed autonomously.The performance of our engineered tools will be analyzed in terms of repeatability, accuracy, maximum stress, strain and load in operation. Robotics and automation technologies will be evaluated in terms of task performance efficiency (e.g., in aeration to identify and break up larvae concentrations), whereas monitoring techniques will be evaluated in terms of their ability to lead to information that helps provide a stable growing environment. Data from all the different sensors will be analyzed using standard-of-practice computer vision and machine perception techniques, whereas our developed algorithms efficacy will be tested against relevant state-of-the-art methods as they become available during the course of the project.Objective 2 milestonesDetermine automation feasibility and efficacy relative to manual efforts,Determine costs of automation relative to manual operations, and areas for further improvement.Manual tending of bioreactors will be considered the baseline condition and occur in parallel with operation of the automated bioreactor system. Response variables will be converted to differences between baseline and automated (experimental) systems. This limits confounding effects of environmental variation. Individual BR2 bins in each system will be considered as "plots" within each system for each experiment iteration. Experiment iterations will be considered as blocks. Each BR2 plot in the manual system will have a corresponding BR2 plot of the same age/handling status in the automated system, and data will consist of the difference in performance between paired manual and automated BR2s. These data will be analyzed using general linear mixed models. We will also track cost inputs for each system (equipment, labor, energy) and estimate the value of outputs (fertilizer, larvae, savings on waste removal) to perform a partial cost benefit analysis and estimation of time to recoup automation investments.Objective 3 milestonesQuantification of the degree of fluctuation in larval, adult, and frass outputs over time relative to waste inputs to estimate stabilityDemonstration of continuous rearing on heterogeneous farm waste inputsSystem success will be indicated by completion of at least three generations, and failure will be indicated by bioreactor "crashes" that result in loss of the rearing operation. Stakeholder engagement and perceptions of automation in BSF rearing will be evaluated through surveys