Source: SAFETYSPECT, INC. submitted to
COVID-19 RAPID RESPONSE: INSPECTION AND DISINFECTION SYSTEM FOR MANAGEMENT OF COVID-19 IMPACT ON FOOD FACILITIES
Sponsoring Institution
National Institute of Food and Agriculture
Project Status
TERMINATED
Funding Source
Reporting Frequency
Annual
Accession No.
1023900
Grant No.
2020-33610-32479
Project No.
CALW-2020-06082
Proposal No.
2020-06082
Multistate No.
(N/A)
Program Code
8.5
Project Start Date
Sep 1, 2020
Project End Date
Apr 30, 2021
Grant Year
2020
Project Director
Vasefi, F.
Recipient Organization
SAFETYSPECT, INC.
10100 SANTA MONICA BOULEVARD SUITE 300
LOS ANGELES,CA 900674107
Performing Department
SafetySpect
Non Technical Summary
The contamination and sanitation inspection and disinfection (CSI-D) system developed under thisproposal is in response to the food safety subsection of the special COVID-19 USDA NIFA SBIRProgram. SafetySpect proposes developing a fast, convenient, and easy to use the handheld system forinfection prevention related to COVID-19 in food preparation and serving facilities. Novel aspectsof the CSI-D solution include the combination of contamination identification and immediateremediation of the potential threat (bacteria, virus) using UVC light disinfection, and documenting this process to provide traceable evidence of disinfection. This system provides evidence for follow-on investigations, such as tracing cross-contamination and tracking sanitation, in addition to trainingcleaning crews, and assessing and improving processes and sanitation standard operatingprocedures (SSOPs). SafetySpect's new technology will improve safety for food processing staff,cleaning staff, as well as food service customers.The objectives of this SBIR Phase I proposal are: (i) Design and build the CSI-D prototype andsoftware, (ii)Validate the illumination system uniformity and effectiveness of viral disinfection, and(iii) Perform an optical safety assessment and develop personnel UV safety gear requirements andsanitation procedures. During this project, SafetySpect will also analyze eye/skin safety issuesrelated to UVC light exposure and create protocols for operating the system safely.This technology targets the institutional and restaurant food service industry. The restaurantindustry market was mature before the COVID-19 pandemic hit. Now the restaurant industry needsto adapt to current safety/environmental protocols in the hope to return to previous sales volume.The goal of SafetySpect is to provide this industry with a means to reassure their customers andstaff that they can operate safely in this new reality.
Animal Health Component
50%
Research Effort Categories
Basic
(N/A)
Applied
50%
Developmental
50%
Classification

Knowledge Area (KA)Subject of Investigation (SOI)Field of Science (FOS)Percent
50340301000100%
Goals / Objectives
The overall objective of this project is to improve workplace safety for food production and food service workers, in addition to providing a safe environment and food service for customers. All FSIS-regulated establishments are required to have Sanitation Standard Operating Procedures (SSOPs), which are written procedures that an establishment develops and implements to prevent direct contamination or adulteration of product. It is the establishment's responsibility to implement the procedures as written in the Sanitation SOPs. The establishment must maintain daily records sufficient to document the implementation and monitoring of the Sanitation SOPs and any corrective action taken. FSIS verifies that regulated establishments adhere to the procedures in place, and FSIS states that the "sanitary procedures that establishments are already following to protect food safety will also help prevent the spread of respiratory illnesses like COVID-19". There is a need to document the implementation and monitoring of these SSOPs as they are adapted for COVID-19. This protects cleaning staff and ensures safety during inspection and sanitization in addition to allowing for assessing and improving processes and sanitization standard operating procedures (SSOPs). The CSI-D technology includes a means of recording, documenting by recorded images and video. CSI-D can wirelessly communicate the inspection process so that remotely located personnel can provide oversight and respond to inspection issues immediately.SafetySpect goal is toidentify the following opportunities for CSI-D: 1) visual documentation for monitoring the cleanliness of human contact surfaces, 2) a non-contact disinfection method using UVC that is safe on surfaces while eliminating unnecessary aerosol distribution of unpleasant chemicals, 3) a technology that can be integrated with current SSOPs while addressing COVID-19 issues and 4) modular software applications specifically designed for the food industry that direct the inspector user experience (e.g., checklists) and ensure best practices are achieved.SafetySpect's CSI-D system competitive advantages are: Portability; Low Cost; Instant contamination detection; Immediate disinfection; Photographic record; Easy SSOP integration; Proprietary technology.
Project Methods
The main components required to develop the CSI-D system for U.V. wavelengths that are suitable for saliva and respiratory droplet detection have been specified and are available from stock or with short lead times. SafetySpect has commenced the acquisition/sourcing of specific potential components for our CSI-D proof of concept research. These components include: 1) LEDs UVC 280nm, 2) U.V. sensitive camera, 3) U.V. lens, 4) band-pass filter, and 5) lithium batteries. We have a ZEMAX model of the UVC LED and imaging system for illumination design and optimization. We will use a system-on-chip module connected to the camera to control the system and communicate wirelessly with the tablet (for operator guidance and data management).We will deploy the software in a cloud-connected tablet app which will manage data flow for inspectors and management staff with an appropriate data access hierarchy. The system will store and document contamination inspection, remediation through cleaning and sanitation, and provide traceability of sanitation for the airline industry as well as regulatory bodies.To provide a simple and fast solution for surface inspection, the image capture and assessment must occur automatically as the operator moves through the food serving or processing facilities. For each point of inspection, the system captures two images: one with LED "on" (fluorescence image) and one with LED off (background). The steps involving image processing include:Pre-processing: There are about 50ms between the background image (imaged captured before LED on) and the actual image (image captured after LED on). Image registration methods will be applied (if necessary) to the two images to correct for possible motion artifacts. By subtracting the background image and the actual image, a "calibrated image" will be generated.Image Segmentation: Image segmentation techniques, such as Otsu thresholding K-means, will be used to segment the calibrated image and extract hotspots and blobs within the image.Hotspot Feature Extraction: Shape analysis will be performed to extract textural/shape features, such as area. We will also extract statistical features (mean, median, interquartile range) as a measure of pixel intensity.Decision Making: After the analysis of shapes, a final decision will be made whether droplets are detected or not. Based on the features discovered in the hotspot feature extraction step, we will investigate statistical (principal component analysis), clustering (K-means, Fuzzy C-means), and rule-based approaches for droplet binary detection. In the rule-based approach, multiple thresholds will be used to eliminate false positives. Adaptive thresholds will be calculated to avoid the impact of ambient light, and the device/scene response variability.Validation of viral disinfection is an established procedure. It involves spraying a surface with a solution containing a known concentration of virus and then letting the surface dry for an appropriate period of time for the particular virus model. The surface then has disinfectant applied to a specific region. After the disinfection, samples are collected by wiping the disinfected area and the nondisinfected area and analyzing the samples for viral load. For our application test, instead of applying a liquid disinfectant, we will expose an area on the test surface with CSI-D UVC illumination (6 inches in diameter).We will use two methods to quantify UVC exposure during the device use:UVC instant power measurement using U.V. optical power meter at various distances and angles from the device, reflection from inspected surfaces and nearby uninspected surfaces such as walls or floors.UVC cumulative dose measurements will be assessed by UVC dosimeter cards attached at different points on the operator's body and face-shield to measure how much UVC dose the operator will receive during a full inspection job.We will investigate the effect of using various directional light shields such as hoods and shades around the hand-held device and other barriers such as portable shields and drapes in the area of operation. This will help us develop safety protocols and training materials for various operational scenarios.

Progress 09/01/20 to 04/30/21

Outputs
Target Audience:This product targets the institutional and restaurant food processing and service industry. The restaurant industry market was mature before theCOVID-19 pandemic hit. Now the restaurant industry needs to adapt to current safety/environmental protocols in the hope of returning to previous sales volume. Some market facts from the 2019 Restaurant Industry Factbook of The National Restaurant Association include: • Restaurant sales in 2019 were $863 billion • 4% of 2019 US GDP comes from the restaurant industry. • 12.6% of the American family budget is spent on food, and 43% of that goes to restaurants • 16.9 million restaurant employees were expected by 2029. • 33% of Americans got their first job experience in a restaurant In addition to market size, it is important to understand that the restaurant industry has a high propensity to adopt new technologies, particularly those that can reduce costs, increase revenues, or increase efficiency: • 95% of restaurant owners observed an increase in their business efficiency by embedding technology in their process • 73% of guests believe that technology adds more to their dining experience • 80% of restaurants are utilizing various types of technologies to aid them in running their business more efficiently (online reservations, ordering, inventory, and analytics) Market drivers for CSI-D are a combination of anticipated enhanced regulation, reputation protection, and customer reassurance. Post COVID-19, restaurants will need to convince customers that their food service environment is safe to continue to grow revenue. Educational food services will need to convince parents, state and local authorities, staff, and students that their environments are safe. Restaurant chains are already experimenting with new self-service order methods, restaurant layouts, sanitation protocols, and enabling technologies. There is a high probability restaurant and institutional food service will adopt technologies that can reliably address COVID-19 as well as bacterial and viral infection threats, and that can be highlighted to stakeholders to increase confidence. Changes/Problems:We encountered several problems during design and prototyping, these include optical and electrical issues. We had a problem with EMI noise generated by LED drivers during LED flashing that interfered with SOM board operation and disrupted normal imaging and data transfer operation. We had to developed shielding methods for boards, wires, and connectors to overcome the EMI issue. With respect to optical issues, we have encountered problems with our fused silica optical window internal reflections and optical filtering artifacts. The fused silica window had antireflection coating but still reflects enough illumination light leakage from multiple reflections and through our optical filters even though they have >OD4 blocking. To get around this we are considering adding additional filtering, but our primary solution is to replace the optical window with a perforated metal plate with apertures in front of LEDs and cameras. This also significantly reduce the cost of the device (the optical glass costs about $500) and provide additional robustness by avoiding front glass breakage. The other issue we encountered is due to the wide-angle view of one of the camera lenses which creates a shadowing effect at the image periphery due to the more oblique angle of incidence. We have two solutions for this, one is to use a lens with smaller field of view and the other is to digitally mask the image periphery during display. What opportunities for training and professional development has the project provided?Commercialization Successes: USDA NIFA SBIR Phase I funding has helped us to secure additional funding to further develop the commercialization of the CSI-D technology. After securing additional private investment, we have selected and contracted ComDel Innovation (https://comdelinnovation.com/ , Wahpeton, ND) as a contract manufacturer for our assembly of the CSI-D system. We have begun the production of commercial beta prototypes for laboratory and field testing and pilot studies. We anticipate having up to 30 beta prototypes of the CSI-D system by the end of May 2021. We are planning to deploy up to five CSI-D beta units in the pilot studies during this SBIR Phase II project. We expect to sell or lease up to 20 CSI-D beta units in 2021 to universities, government laboratories, and early adopters. We will keep five units for backups and emergency replacements. We have begun implementing a quality system that meets FDA 21 CFR 820 and ISO13485 requirements. We are implementing an electronic quality management system using Qualio (https://www.qualio.com/) and we are in the process of transferring all paper-based SOPs and forms into the electronic system. We have been updating our financial model for potential investors and have built a due diligence repository of financial documents and business plans necessary for venture capital and family office investors. We have revised our website and marketing materials. We will be using this to approach potential customers, strategic partners, investors and to expand awareness of the future potential of our solutions for other industries. We have been working on our public relations strategy and have had a number of articles appear in the newspapers and online highlighting our collaborative work with the University of North Dakota. Based on the results from Phase I, SafetySpect has approached multiple food processing businesses and commercial kitchen facilities regarding pilot studies. We have received support letters from five of these, reflecting their interest in the CSI-D technology and their willingness to be our first customers upon the success of pilot studies in their facilities. How have the results been disseminated to communities of interest?We are drafting one journal publication with USDA-ARS scientists and UND researchers describing CSI-D+ technology and novel features. What do you plan to do during the next reporting period to accomplish the goals? Nothing Reported

Impacts
What was accomplished under these goals? During Phase I, SafetySpect developed contamination, sanitation inspection, and disinfection (CSI-D) system and brought the hardware and software maturity to TRL 6-7, where we can demonstrate an actual system prototype in an operational environment. The CSI-D system is a fast, convenient, and easy-to-use handheld system. The CSI-D solution includes contamination identification and immediate remediation of potential threats (bacteria, virus, and fungus) using UVC light disinfection, and documenting this process to provide traceable evidence of disinfection. We aim to provide evidence for follow-on investigations, such as tracing cross-contamination and tracking sanitation, in addition to training cleaning crews, and assessing and improving processes and sanitation standard operating procedures (SSOPs). In Phase I, SafetySpect promised to develop a prototype for the detection of saliva and respiratory droplets on surfaces and validate the imaging and UVC disinfection capabilities of the system. We also promised to complete an optical safety analysis and to incorporate safety features such as UV safety gear requirements. Not only, have we achieved the targeted milestones, but we also accomplished further development of the system, software, and commercialization milestones beyond the original project scope which will be described below. We have completed the optical design, mechanical design, heat management, electrical design for illumination and data processing and built two CSI-D minimum viable product which can be used in field pilot studies. We have developed the firmware for the system on a module board that controls illumination and image capture and incorporated an Android smartphone as a smart display and secondary processing system. The smartphone also provides network connectivity to a cloud server which hosts a management dashboard for task management and data storage and display. Further accomplishments beyond original scope: Based on user discovery we determined that it was essential to provide an inspection tool that not only detects saliva and respiratory droplets for COVID-19 response, but also detects other organic residues such as fecal matter, and bacterial contamination. Because of this our prototypes include both 275 nm LEDs for saliva detection and germicidal features, and 405 nm LEDs for other organic residues and bacteria. We also integrated two cameras and two optically filtered imaging paths to collect multiple fluorescence emission wavelengths. We completed the illumination system design to ensure adequate illumination uniformity and intensity at desired working distance (between 5 to 8 inches) for fluorescence detection and for effective disinfection within 2-5 seconds. We have tested effectiveness of bacterial and viral disinfection in laboratory environment at the University of North Dakota. We have characterized killing percentage based on multiple UVC dosages (changing LED power, distance, and time) for theStreptococcus pneumoniaebacteria and influenzas H1N1 virus models. Further accomplishments beyond original scope, we have tested UVC effectiveness on Aspergillus fumigatus spores. We are preparing a publication based on the germicidal effectiveness results and contamination detection results on multiple organic residues and saliva.

Publications