top of page

Seeing The Light - How Spectral Analysis Is Changing Food Quality Testing

Written by Lisa Johnson & Mohammed Kamruzzaman

Lisa Johnson serves the fresh produce industry as a Consultant with AgTech Insight, and as an advisor to InSight Labs, a company improving produce quality control at the speed of light!

In this article, we explore the revolutionary power of advanced computing by looking at its impact on spectroscopy and AI as applied to quality control within the fresh food processing industry. Never has it been more critical to accurately and efficiently assess food quality in the wake of the COVID pandemic and changes throughout not only our US food supply chain, but the entire world food system.

By applying spectral analysis, fresh fruits and vegetables can be analyzed based on customer driven parameters with 95% or higher accuracy. The critical innovation, up until now unavailable to spectroscopy equipment manufacturers, is the use of machine learning algorithms and chemometrics to build a reliable library of spectral signatures. Once established, these spectral signatures can be used to measure characteristics like pH, shelf-life, viscosity and sugar content. This represents a sea change in the industry where traditional methods have failed to produce a fast, reliable test.  The authors also cover the triple bottom line benefits of this new SAAS (Software as a Service) methodology covering and its implications for not only profit, but also people and the planet.

Seeing the Light - How Spectral Analysis is Changing Food Quality Testing


Lisa Johnson and Mohammed Kamruzzaman


Innovation in ag tech is advancing rapidly, radically changing the requirements for food quality control. Food processors have a high demand for precision quality measurements, but are located in the “messy middle” of the food supply chain where few technologists have dared to tread. Despite this, they still require transparency, traceability and accuracy in their measurements and analysis across multiple product lines. Likewise, sustainability and safety are becoming increasingly important across the entire spectrum of food production and reducing

waste, resource usage, and potential harm to humans and the environment is paramount.


When it comes to quality control, the industry needs a fast, accurate, simple and reliable solution. In the food processing portion of the food supply chain for instance, thousands of sample tests are done every day to identify various traits affecting food quality such as; freshness, pH, nutrient content and chemical compositions.  Unfortunately, traditional methods can lead to errors, processing delays, and waste due to a lack of innovation and modernization of the tools necessary for more accurate testing.  What has been needed for a long time in food quality

analysis is a product with a higher level of efficiency and effectiveness.


Any new technology introduced to the market today should also have a triple bottom line impact. This new, high bar has helped drive not only revenue efficiency, but also improved outcomes in the social and environmental indicators as well. Spectroscopy offers the possibility of a paradigm shift in the food processing industry. Recent advancements in machine learning (ML) and

artificial intelligence (AI) due combined with an exponential increase in computer processing power, have made it possible to make the kind of dramatic gains necessary for highly accurate quality control platform tools.

Benefits of combining spectroscopy with AI

The key drivers for new technology are product-solution and product-market fit. This synergy then creates a match that ultimately has a positive impact on the bottom line. (Ideally, the innovation will also help drive social or environmental impact) Technology that combines spectroscopy with AI is capable of creating the triple bottom line impact the industry needs.

Creating a software-as-a-service (SAAS) offering by combining a stepped approach to AI that works in harmony with spectral analysis, can deliver fast, effective measurement results. An optical analysis subscription service should be; flexible enough to work with any type of produce, agnostic to the type of food processing equipment, and able to accurately measure the input quality of end products. (I.e. ketchup, granola bars, etc)


The combination of spectroscopy + AI can decrease revenue loss and increase capacity in a few ways:

  • First, it saves time by streamlining processes; one measurement with spectroscopy replaces a wide variety of mechanical, physical, or chemical tests, bringing analysis from up to 36 hours, down to minutes. This delta not only saves valuable time and money, but increases operational efficiency.

  • Secondly, no expensive reagent is necessary and little to no sample preparation is needed while the descriptions of salts, sugars, fats, protein, water content, viscosity, and more, are available with one measurement, saving money on lab supplies and equipment.

  • Third, data can be collected in one place, and results aggregated into an easy-to-read dashboard, enabling quick decision-making with a platform for in-depth performance analysis over time.

This change in testing can democratize the data and make it available to more employees across the organization, increases transparency and builds a culture of continuous process improvement.

Meeting benchmarks and making strides is possible when the team works together, and employees can be driven to improve by allowing feedback and ideas across all levels of the enterprise. Food quality degrades over time. Therefore, if decisions are delayed while the QC (quality control) function takes hours or days to complete testing and analysis, productivity loss can be staggering, into the hundreds of millions of dollars. Reducing time, reagents and equipment, errors, data entry and analysis, is now a possibility for food processors with a new, hardware-agnostic platform that offers optical analysis across all light frequencies. Reuse of data is a key element in efficient platforms.  Ideally, this new software platform can be built on a client-centric spectral library comprised of data sets which enable rapid on-boarding for targeted customers and accurate results. This is the revolution that will replace chemical, mechanical, and qualitative analysis in the food industry for not only food processors, but the entire food supply chain.

The Modern QC Lab

A wide variety of techniques and equipment are used in QC labs to determine if a batch of production will meet standard parameters for quality. These steps are layered with complicated protocols that include several reagents and stressful steps to determine the chemical attributes of food samples. Food processing plants still use wet bench procedures along with a variety of bulky analytical equipment for measurements making data collection difficult and disjointed.


Food processors are realizing the value data across their supply chain, and the need for a new analytical standard to optimize productivity. The ideal solution, for example, uses a variety of portable spectrometers to determine chemical compositions without the negatives of wet bench procedures allowing easy accommodation in the QC lab or for the instrument to be taken to the sample if necessary. Although the interface can accept data from a wide variety of spectroscopy instrumentation, this portability allows for use throughout the manufacturing process for monitoring quality control. Portable spectroscopy instruments can be used to inspect samples during production, allowing processors to decide in real time whether to reject/accept a batch, real-time process monitoring, process optimization and end product inspection.


Typically, labs today run several procedures through different instruments, which are not harmonized or synchronous. Data entry is manual, in order to aggregate information among several parameters. Introducing this human element can create errors leading to lost time and money.

To reliably guarantee accurate results, the industry needs to develop a deep learning model with a well-defined user interface that properly takes into consideration food processing workflow to ensure minimal interruption on their core operational processes which is required to guarantee accurate results. This type of platform will enable each quality parameter to be tracked to provide a deeper dive into the samples, while the visual representations make results straightforward and transparent which makes decision-making for process change or waste control quick and easy. The industry needs accurate spectral analysis which is why the AI models need to be in constant iteration as they are created from routine sampling of huge numbers of food samples, representing a wide variation in and among products. These models are optimized and validated through automated calibration updates as the platform “learns” more and more which enables real time decision making for adjustments to conditions in processing plants.

A new generation of quality control in food processing looks totally different

What was impossible a few years ago is now made possible by bringing together advances in several areas of research and technology with the beneficiary being the food quality and safety industry and its triple bottom line. The modern food quality control lab is mobile and generates data in seconds. Reducing revenue loss, food waste and processing time comes neatly packaged in a user-friendly platform, ready to inform critical decision-making. Since the outbreak of COVID-19 on a worldwide scale, many stakeholders in the food industries are rethinking their methods, procedures and operations to develop a much more resilient food supply chain. It’s more than saving time and money, it’s about reduction in greenhouse gases and preserving a pristine environment for future generations.


Lisa Johnson serves the fresh produce industry as a Consultant with AgTech Insight, and as an advisor to InSight Labs, a company improving produce quality control at the speed of light! She also serves as an Adjunct Assistant Professor at North Carolina State University. Lisa is committed to finding ways to reduce food loss and waste and her focus is on finding solutions that incentivize or somehow benefit growers. Check out her profile on our Team page or visit her website at!

bottom of page