Our Science
Our Technology Lets You See The Invisible
SafetySpect combines proprietary hardware, AI, and machine learning to create meaningful insights across industries
-
All matter absorbs and reflects photons of different wavelengths (colors) at a molecular level in a way that is reproduceable and identifiable. We call these spectral signatures, similar to a multi colored barcode.
-
Our sensors detect multiple wavelength bands simultaneously to look for these unique spectral signatures in real time.
-
Spectral signatures analyzed with advanced machine learning techniques provide valuable information about the biological and molecular composition of surfaces, far beyond human perception.
-
The recent advent of powerful miniaturized components including light sources, sensors & power sources, combined with device level edge computing, enables handheld systems with unique capabilities that would not have been possible in the past.
Multispectral Fluorescence Imaging (MFI)
Making Invisible Contamination Visible
CSI Available Now!
CSI Technology
The CSI (Contamination, Sanitization, Inspection) handheld system is SafetySpect’s first commercial product, designed specifically to detect and display invisible contamination in real time for immediate documentation and action.
CSI Innovation
-
The first to use multiple cameras and wavelength bands for more accurate results.
-
The first to detect fluorescence in ambient lighting conditions, thanks to our proprietary pulsed illumination and detection protocols.
-
Advanced spectral libraries and AI with machine learning to identify contamination such as blood, fat, biofilms and feces.
How it Works
A number of biological molecules emit light when excited with the proper wavelength, without the addition of any reagents or chemicals.
Some of these molecules are present in all living things (NADH, Tryptophan, Phenylalanine) others are specific to certain organisms, such as chlorophyll, fats and oils, and other pigments produced by plants, animals, fungi and bacteria.
The fluorescence emission of a complex sample like a biofilm will be a combination of the fluorescence of all its components at their specific ratios, creating a fluorescent barcode unique to that sample.
Software and Data Processing
-
Onboard software is fully customized to your workflow and application and includes everything you need to acquire, display, analyze and act on the data in real time. Features include facility maps and task checklists, automatic timestamped data recording, supervisor approval processes, and more.
-
After synching, the web dashboard allows managers to review data and trends for long-term decision making and auditing, increasing accountability across the organization.
-
Machine learning models identify long-term trends, cross-contamination sources and other workflow issues.
-
Ensuring the privacy of our clients is a key concern for us, which is why we use federated learning and implement a decentralized system to secure sensitive information. We are currently working toward FedRAMP compliance so you can be assured your information remains secure.
Selected Publications
Deep learning and multiwavelength fluorescence imaging for cleanliness assessment and disinfection in food services
Frontiers in Remote Sensing, 2022.
Combining deep learning and fluorescence imaging to automatically identify fecal contamination on meat carcasses
Nature Scientific Reports, 2022.
MultiMode Spectroscopy System (MSS)
Powerful Point-Scanning and Analysis in the Palm of Your Hand
-
The point-scanning Multimode Spectroscopy System (MSS) combines the power of multispectral fluorescence with reflectance spectroscopy in the visible + near-infrared (VIS-NIR) and short-wave IR (SWIR)
-
No other hand-held instruments on the market combine these three powerful modes.
-
More modes = more diagnostic power
- Three modes probe different molecular characteristics.
- What’s discernible in one mode may be impossible to determine in another.
SafetySpect’s Quality, Adulteration and Traceability (QAT) Device
-
Unlike MFI technology, reflectance does not require absorption and emission of a photon. Reflectance is what our eyes detect every day in the visible (400-700 nm) region of the electromagnetic spectrum.
-
Similar to fluorescence, each material has its own reflectance spectrum.
-
Combining multispectral fluorescence, VIS-NIR and SWIR reflectance modes gives us a larger spectral barcode to detect very subtle changes in composition
-
Our powerful onboard machine learning algorithms are able to identify not only the composition of a material, but also factors such as species and freshness, often without removing it from the packaging.
-
Our first product in this space is the Quality Adulteration and Traceability (QAT) device, designed to monitor food quality and freshness.
Selected Publications
Freshness in salmon by hand-held devices: Methods in feature selection and data fusion for spectroscopy
ACS Food Sci. Technol. 2024.