Source: FIELD DATA SERVICES LLC submitted to NRP
AUTOMATED POLLINATOR IDENTIFICATION: NEXT-GENERATION SENSORS FOR SPECIES AND BEHAVIOR
Sponsoring Institution
National Institute of Food and Agriculture
Project Status
COMPLETE
Funding Source
Reporting Frequency
Annual
Accession No.
1028623
Grant No.
2022-33530-37405
Cumulative Award Amt.
$165,685.00
Proposal No.
2022-01412
Multistate No.
(N/A)
Project Start Date
Jul 1, 2022
Project End Date
Feb 28, 2023
Grant Year
2022
Program Code
[8.13]- Plant Production and Protection-Engineering
Recipient Organization
FIELD DATA SERVICES LLC
454 PARMA DR
ESSEX,MT 599169730
Performing Department
(N/A)
Non Technical Summary
Producers need to measure pollinator population size, distribution and effectiveness in numerous remote locations. Ecosystems involving commercial bees, feral bees and native pollinators along with weather, climate and other environmental factors are complex. Pollination effectiveness within these complexities can be a key component of crop yields and profitability. Understanding this complexity requires new tools for producers that reveal which species are present, in what abundance, and which specific portions of cropland could benefit from modified pollinator management practices. Beyond the immediate needs of producers are the educators, students, citizen scientists and agency land managers who have all taken great interest in pollinator health and what they can do as individuals to enhance the habitats and systems that support pollinators. This Phase I SBIR project will determine the technical feasibility of using a newly available LiDAR on a chip (LoC) to detect insect pollinators and characterize pollination events in the field.Numerous teams are working hard to apply AI-powered machine vision technologies to bear on identifying pollinators and other wildlife. The problem with these technology-intensive devices is that they are expensive, require huge batteries and high bandwidth internet connectivity. The way to make AI-powered devices affordable and practical for field use is to use a micro-powered motion detector to turn on the AI power only after a moving target is detected. Commercial trail cameras use a passive infrared (PIR) sensor for this purpose. The problem with a PIR sensor is that it will trigger only if the moving object is of a different temperature than the background, like a large warm-blooded animal moving across a cooler background. In October 2020 STMicroelectronics introduced the first commercialized multi-zoned rangefinder which is essentially a small LiDAR on a tiny $9 computer chip. This sensor does not require a temperature differential to detect motion. Our team will be testing the ability of this new sensor to detect small moving objects under a variety of outdoor conditions as a way to produce a small, affordable, micro-powered device for detecting and monitoring pollination events in the field.
Animal Health Component
90%
Research Effort Categories
Basic
5%
Applied
90%
Developmental
5%
Classification

Knowledge Area (KA)Subject of Investigation (SOI)Field of Science (FOS)Percent
21172991130100%
Goals / Objectives
This Phase I SBIR project will determine the technical feasibility of using a newly available LiDAR on a chip (LoC) to detect insect pollinators in the field. There are substantial commercial and ecological needs to detect, quantify and characterize the effectiveness of pollinators on crops and critical wild plants. Current time-consuming methods that rely on manual observations are being augmented by powerful new machine vision, machine learning and other AI powered technologies. As promising as these new image-based technologies are, they are limited by the cost, complexity and electrical power required to bring the power of AI-enhanced computer vision to remotely located agricultural lands.If a producer wants to respond to a pollination deficit in time to preserve adequate yields, the producer needs to know the current state of pollination in almost real time during critically short blooms. Ineffective commercial hives can be mitigated on the spot, and building a record of year-to-year effectiveness of commercial, feral and native pollinators can help formulate site-specific pollinator management plans that account for weather and local pollinator habitat conditions.Leveraging rapidly evolving computer vision technologies to meet these producer needs will require new low-cost and low-power ways to bring these computer technologies to all corners of remote croplands. Running power-hungry computer vision cameras full-time in the field will fail to meet the logistical constraints of real world food production. Commercial camera traps, aka trail cameras, are designed to acquire images of critical events without the need for huge batteries, solar panels and cloud connectivity. Camera traps achieve this by leaving the power-hungry image sensor turned off most of the time. Only one small micro-powered electrical component is left turned on and searching for movement. This small component is a passive infrared (PIR) sensor. Only after the PIR sensor detects motion is the power-hungry image sensor turned on to acquire images. After the images are acquired and stored the power-hungry image sensor and main processor are again turned off to save battery power.A similar approach will be taken by this project in order to increase the feasibility of using computer vision technologies for monitoring pollinators. Micro-powered PIR sensors that are used in trail cameras would solve the power supply issue except for the fact that PIR sensors are only triggered when the moving target presents a temperature differential with a static background. Camera traps work well on large warm-blooded animals moving against a cooler background but PIR sensors are not always effective for detecting flying insects. PIR sensors are also susceptible to false triggers from sunlight reflecting off of vegetation that is moving in the wind.Another way to detect a moving object is using the image sensor enhanced by onboard digital imaging processing techniques. Onboard image processing can acquire a series of baseline images to determine if a new object has moved into the field of view. This process is called background subtraction because when a new image is taken, the pixels are subtracted from the previously acquired baseline images, pixel by pixel, to identify clusters of pixels that are new and different from the baseline. This method of detecting moving objects does not require a temperature differential of the moving target, but it has two disadvantages: 1) Image background subtraction requires that the power-hungry image sensor be turned on almost continuously, and 2) background subtraction is susceptible to false triggers from background vegetation moving in the wind.This project will evaluate a new and inexpensive electrical component for detecting insect pollinators that might replace or augment PIRs and image background subtraction for achieving cost effective monitoring of remote pollination events. In late 2020 STMicroelectronics introduced the first commercialized multi-zone rangefinder on a chip. This micro-powered $9 chip produces an 8x8 grid of LiDAR distance measurements. This grid of accurate distance measurements is achieved by emitting an 8x8 grid of tiny invisible infrared laser pulses. The chip measures the time between emitting a pulse and the return of reflected pulses. From these measurements the chip accurately measures the distance (within a couple of millimeters) to the nearest object within each of the 64 zones in its field of view.This LoC sensor has a field of view of about 45 degrees, and it can measure distances of up to several meters. This field of view and measurement capability is well-suited for detecting medium to large insect pollinators. An LoC sensor uses slightly more battery power than a PIR sensor but it uses orders-of-magnitude less battery power than operating an image sensor full time for background subtraction and image object recognition.The enticing potential of leveraging LoCs for monitoring pollinators begs further investigation. However, being a new and undocumented application for LoCs our team must first conduct substantial feasibility testing. The goal of this Phase I project is to measure and determine the envelope of conditions for which LoCs can be used to solve the problem of cost-effective use of machine vision technologies for monitoring insect pollination of agricultural crops and critical wild plants.To determine the envelope of conditions under which LoCs will be effective. An automated test system will be built that will automatically move variously sized targets past through an LoC sensor's field of view. The automated test systems will compile extensive and precise tabulations of positive detections and false negative and false positives. These test data will be taken while modifying target reflectivity, speed, size and distance from the sensor. Background conditions will also be manipulated and tested including moving vegetation and over a range of realistic outdoor lighting conditions.The goal of these tests will be to narrow and carefully define the range of insect characteristics and conditions for which LoCs will be effective. From this information an optimized device can be designed and evaluated for commercial potential and Phase II project feasibility. The following objectives were developed to meet the project goals:1) Identify and mitigate crosstalk noise that might affect LoC performance. This objective involves determination of electronic noise impacts on LoC performance from other electronic subsystems in a smart camera system such as high-speed clocks, USB, GPS and main processor.2) Measure and mitigate unit to unit variation of LoC sensors. LoC sensors being a newly commercialized component it is prudent to look for unit-to-unit variation, and if found, mitigate it by screening components or other measures.3) Experimentally verify that insect detection accuracy with LoC or an LoC plus smart camera (LoC+Cam) system is as good or better than existing machine vision devices.4) Experimentally verify that honey bee sized insect pollinators can be reliably detected when within a specified range of size, flying speeds and distances.5) Experimentally verify that software analysis of LoC and LoC+Cam sensor data can distinguish between major insect taxa.6) Experimentally verify end-to-end system performance of battery, data acquisition and analysis appropriate for remote field deployments.7) Experimentally verify that software analysis of LoC and LoC+Cam sensor data can detect insect pollination events.8)Resources permitting, we can address two stretch goals: A) determine whether an array of LoCs is substantially better than a single LoC unit. And B) determine if LoC assisted focus of a high-end camera board can improve detection of species and behaviors.
Project Methods
Technical feasibility testing will be narrowly and intensely focused on characterizing the ability of a LiDAR on a chip (LoC) to detect and monitor insect pollination in the field. Toward this end, a computer controlled test system will be designed and constructed that will move a simulated insect across the LoC's field of view at various speeds and distances. The computer will compile numerous instances of detections, false negatives and false positives under the varying conditions. Conditions that will be varied will include the size, reflectivity and wing orientation of the simulated insects, the speed and distance that the target moves across the field of view, the lighting conditions, and the type and movement of the background. The compiled results will be analyzed using common Gauge R&R statistical methods to determine the reliability of LoCs at detecting insect pollinators under the various conditions. Results from the following test and measurement efforts will be evaluated for their impacts on product design, partner outreach and initial commercialization strategies:1)System level electrical crosstalk that might degrade LoC sensor performance.2)LoC unit to unit variation that might require mitigation to achieve reliable product performance.3)LoC detectability of three categories of pollinator sizes: honey bees, bumble bees and moths/butterflies.4)End to end performance of LoC and LoC plus smart camera systems for detecting pollination events in terms of field robustness, battery power consumption and detection efficacy.5)Comparison of data usefulness between devices that employ only LoCs, LoCs plus a PIR, and LoCs plus a smart camera.The goal of these experiments will be to de-risk a commercial launch in 2023-2024. De-risking commercialization requires optimizing the design and target applications armed with a detailed understanding of the technical capability of LoCs to detect insects. Risk associated with commercialization can be minimized by targeting the mechanical design of the device for optimized distances to targets, sun shading and orientation. Detection efficacy can be optimized by developing the firmware that will extract insect detections from background noise. Optimizing and understanding the product's design and constraints will inform our team choice of field applications with the highest chances of success. Determining the envelope of targets and conditions will alter our team's technical directions, target audiences and potential customers.The spreadsheets of statistically analyzed detections will provide a breakdown of size, speed, distance, lighting and backgrounds. From these data we will extract detection reliability at, for instance, 50, 75 and 90% over size, speed, distance and background. From there we can analyze detectability over three broad categories of targets - honeybees, bumble bees and moths/butterflies. Armed with the detectability of these three categories our team can overlay pollinator dependent crops that could support initial beta testing and outreach during early commercialization. Using publicly available data on the market sizes of pollinator dependent crops our team can narrow our focus toward the market segments that will be most conducive to a successful launch. The above discussion of target audiences will also be an important part of the analysis. Outreach efforts with the Xerces Society and the complimentary pollinator projects at Montana State, Penn State, Kansas State, and UC Santa Barbara and UC Davis will all provide valuable inputs into market launch strategies.Overall, the methods and analyses of the LoC technical performance will be straightforward - simple statistical analyses of pass/fail detections over conditions. The automated nature of data acquisition will lend itself to large sample sizes and robust statistical confidence.

Progress 07/01/22 to 02/28/23

Outputs
Target Audience:Our project targetstwo groups. First, producers andgrowers, meaning those people who farm food plants. Second, the researchers who collect information to help producers make decisions. We focused on those who work with insect-pollinated food plants. We are testingtechnology to help producers and agricultural researchers makedecisions about how they manage their land and crops. Management can be very expensive, such as choosing to grow wild insect habitat instead of food, or hauling in commercial bee hives when there are not adequate wild insects. These costs are important, especially for small farmers and economically disadvantaged farmers, who cannot afford to make a mistake in their management and lose profit. That will shut down a producer.For US citizens, these decisions are important because a failure in yields across an industry can lead to an increase in food prices and availability shortages. The nature of the technology is to enable low-power trail cameras for insects. This requires testing alternative technology to traditional cameras to save on battery life and integrating this new technology into an existing USDA-produced platform. 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?We will follow the remaining month-by-month work plan that is outlined in our proposal to USDA.

Impacts
What was accomplished under these goals? Initial software and hardware design are in process. The specific robot hardware is being quoted and specified with sellers. We are developing robot controller software. At this early stage (one month) we are progressing towardgoals 1, 2, 4, and 5. There have been significant developments in software and hardware design that will allow us to move forward with large n-valuetesting of the technology.

Publications


    Progress 07/01/22 to 02/28/23

    Outputs
    Target Audience:Our project targeted two groups. First, producers andgrowers, meaning those people who farm food plants. Second, the researchers who collect information to help producers make decisions. We focused on those who work with insect-pollinated food plants. We tested technology to help producers and agricultural researchers makedecisions about how they manage their land and crops. Management can be very expensive, such as choosing to grow wild insect habitat instead of food, or hauling in commercial bee hives when there are not adequate wild insects. These costs are important, especially for small farmers and economically disadvantaged farmers, who cannot afford to make a mistake in their management and lose profit. That will shut down a producer.For US citizens, these decisions are important because a failure in yields across an industry can lead to an increase in food prices and availability shortages. Changes/Problems:We were unable to work with a live bee hive due to the severe ice storm in Texas. The storm killed off the hives we had arranged to use.Instead, we created a wing-rotating target to study how buzzing wings affect LiDAR vision. This alternative experiment provided us with useful knowledge that we had hoped to gain from the live hive experiment. What opportunities for training and professional development has the project provided? 1. Douglas Bonham has gained multiple coding proficiencies, including proficiency in LabVIEW Professional, and worked with his own custom, and pre-built firmware for the STMicroelectronics LoC chip. Douglas has been in repeated contact with the STMicroelectronics LoC engineering teams, as well as the creator of the OpenMV software that runs on the device's system. Douglas has created new professional networks by gaining insight into the hardware and embedded software needed to integrate the LiDAR into the USDA CIG base device. Additionally, attending interviews with pollination professionals allowedDouglas to increase his knowledge about pollinatorneeds as well as current events in pollinator management. Evelyn assigned multiple academic papers to the team, as knowledge-building activities. Douglas worked closely with the LARTA consultant to gain knowledge about commercialization planning. 2. Constance Woodman gained several new professional skills, including ASCII motor controller coding, and VBA coding. For fabrication, Constance worked closely with onespecialtyfabrication centers to design the salmon simulator fixture,Ragan Ranch Fab Lab,near Dripping Springs, TX. This has increased Constance's capacity for fabrication and engineering design. Additionally, attending interviews with pollinationmanagers about technology needs allowed Constance to increase her knowledge about producerneeds as well as current events in pollinatormanagement. Evelyn assigned multiple academic papers to the team, as knowledge-building activities. Constance worked closely with the LARTA consultant to gain knowledge about commercialization planning. 3.Christopher Evelyn has been included in hardware fabrication, software planning and building, firmware buildings, and machine learning. This has greatly increased his knowledge about custom fabrication, and cutting-edge technology applicable to Ecosystem Sciences. Evelyn is now better able to work on Christopher worked closely with the LARTA consultant to gain knowledge about commercialization planning. Evelyn has created GIS-based maps layers to better understand the problems to be solved and where markets exist, increasing his capacity to work with new kinds of disparate layer data. 4.Ann Bishop has been introduced to Technology Transfer, commercialization processes, granting structures, reporting structures, and technology de-risking approaches. This has increased Ann's ability to relate technology and commercialization to her Marine Biology graduate degree and background, increasing her professional capacity. Bishop has gained skills in specific project management tools, such as Team GANTT, the Microsoft Bug Tracker task tracking format, and various agenda and communication tools. This has increased Ann's ability to vision and plan technology and how it can intersect with ecological sciences. 5. WRichard Hartigan has been introduced to ecological business models to augment his biomedical business background. He has attended multiple pollinator scientist interviews, being exposed to new business and management systems that increase his understanding and capacity in the pollinator technology space. 6. Brian Springer has worked with his first government-funded scientifically based R&D project; he now understands the expectations and process of scientific project development, which works uniquely compared to other R&D or business processes. He is applying skills from his work in finance and procurement to scientific R&D in the Ecology space. This has increased Brian's resume and professional capacity. How have the results been disseminated to communities of interest?We tabled or presented at major conferences, includingEntomological Society of America, Ecological Society of America, Montana Chapter of The Wildlife Society, National Meeting of The Wildlife Society, Columbia Mountain Institute for Applied Ecology. This allowed for hundreds of casual communicationsand building our commercialization hypotheses and producer needs information base. We subsequently conducted long-format interviews of 45 individuals from labs, producers, and applied science firms to better communicate to industry. What do you plan to do during the next reporting period to accomplish the goals? Nothing Reported

    Impacts
    What was accomplished under these goals? We achieved all goals of the project, except for hive rental and live bee testingdue to the state emergency declaration in Texas. The ice storm killed the hives we had lined up to utilize - see the changes section for our alternative experiment. We are most excited by the ability of aLiDAR computer vision alternative to see insect class sizes and behaviors and be able to distinguish between multiple size and behavior classesusing low-power mobile machine-learning architectures. Coming into this project we were sure that low-resolution LiDAR machine vision would be a new way to trigger a traditional camera to reduce the need to always have a camera on. Now we know that the LiDAR itself can provide behavioral data, direction, and size class information without ever having to turn on a camera. This may eliminatethe need for video analysis onboard remote devices, and allow low-power processing of complex data types, with the results beamed directly to the producer over radio connectivity.

    Publications