Applications of spectroscopy in the agri-food industry

posted in: Uncategorized | 0

Applications of spectroscopy in the agri-food industry

This post explores the implementation of a quality detector based on spectroscopy for raw materials in the agri-food industry and agriculture. Here, the use of statistics and artificial intelligence techniques can correlate spectroscopic data with the desired parameters.

Greenhouse. Agri-food and agricultural industry.
Figure 1. Agri-food and agricultural industry. Greenhouse.

In the case of the raw materials processed in the agri-food and agricultural industry, there is a strong relationship between their visual aspect and their quality. In this sense, a rough estimation of their maturity, freshness, or the amount of imperfections can be obtained by a simple visual inspection. However, visual inspection is often not sufficient to guarantee a proper identification and classification. All the previously mentioned factors are of great importance because they influence both in the quality and the price of a product.

Traditionally, the quality sorting process was made by visual inspection. Nevertheless, as it is described in a previous post, human visual inspection is not reliable or as objective as it is commonly believed. Variables like changes in the lighting, the operator who is inspecting the product or the mood of the operator himself/herself can severely influence the process. The unreliability of the inspection is a source of discussion between the clients and the suppliers, whose subjective views usually correspond to their own benefit.

The mentioned problem affects almost every industry in the agri-food and agricultural industry. Here, an objective process where the results do not vary with time, operators or providers is required.

The solution to this problem can be provided by spectroscopy techniques, as it has been described in numerous scientific articles which show positive results in products such as olive oil [1,2], beer [3], wine [4], meat [5] and tomatoes [6].

Products from the agri-food and agricultural industry
Figure 2. Products from the agri-food and agricultural industry that have already been studied with spectroscopy.

Spectroscopic techniques can provide rapid, objective and accurate information for quality sorting competing the conventional time-consuming, expensive and sometimes complex quality processes performed in laboratories, such as FTNIR. In the same manner, usual quality processes use many resources and cannot be applied to the whole batch undertaking the quality process.

In these cases, Pyroistech offers customized spectroscopy solutions for a quick detection of quality parameters as a result of the adequate processing of the spectral characteristics of the light reflected from every sample. This process produces an objective result without the need of trained personnel while making an ideal compromise between visual inspection and laboratory testing.

The principle of the inspection system is the following one:

  • The product is illuminated using a certain light source, such as a COB LED light source or a TAKHI Halogen light source from Pyroistech.
  • The light properties are modified after hitting the sample.
  • The light is collected using a spectrometer or similar apparatus.
  • The obtained information is processed.
  • Processed information permits to obtain an objective and numerical quality result.
Inspection system for industry shown in action. Pyroistech.
Figure 3. Inspection system shown in action that summarizes the above listed points.

For the implementation of the quality process, a methodology based in 3 steps will be conducted by Pyroistech. First, a correlation between the desired parameter (moisture, maturation, color, traditional sorting system etc.) and the spectrum must be obtained. For this reason, a small number of samples is studied with high-performance spectrometers covering ultraviolet, visible and near infrared spectrum.

The spectrum and the parameters measured are the inputs used for the mathematical process where the correlation is made. If the results are optimistic, the second phase will be carried out, that is, a larger amount of samples will be employed to analyze all the casuistry of the particular process under consideration. Different technological approaches are also evaluated in order to obtain the best prototype and assure an optimum performance/cost ratio.

Once the previous study provides successful results, the final product is constructed, which is the third step. A final check of the performance of the equipment in real conditions is required so little adjustments can be made in the algorithms and equipment.

Schema of Pyroistech's workflow in a project for industry
Figure 4. Schema of Pyroistech’s workflow in a project for industry

Finally, it is worth mentioning that spectroscopic techniques can also be implemented in the production line in order to automatically inspect the product directly.  Here, the use of optical fiber can play an important role allowing the equipment to be placed far from the measurement zone, as well as enabling an easy implementation in confined spaces or rough atmospheres (corrosion, high temperature and humidity).


[1] N. Abu-Khalaf and M. Hmidat, “Visible/Near Infrared (VIS/NIR) spectroscopy as an optical sensor for evaluating olive oil quality,” Comput. Electron. Agric., vol. 173, no. April, p. 105445, 2020.

[2] A. G. Mignani et al., “EAT-by-LIGHT: Fiber-optic and micro-optic devices for food quality and safety assessment,” IEEE Sens. J., vol. 8, no. 7, pp. 1342–1354, 2008.

[3] A. G. Mignani and L. Ciaccheri, “Belgian beer mapping and digital fingerprinting using color and turbidity assessment,” Opt. Sens. II, vol. 6189, p. 61892E, 2006.

[4] R. Ríos-Reina, S. M. Azcarate, J. M. Camiña, and R. M. Callejón, “Sensory and spectroscopic characterization of Argentinean wine and balsamic vinegars: A comparative study with European vinegars,” Food Chem., vol. 323, no. April, p. 126791, 2020.

[5] S. Weng et al., “Rapid detection of adulteration of minced beef using Vis/NIR reflectance spectroscopy with multivariate methods,” Spectrochim. Acta – Part A Mol. Biomol. Spectrosc., vol. 230, p. 118005, 2020.

[6] L. Xie, Y. Ying, T. Ying, H. Yu, and X. Fu, “Discrimination of transgenic tomatoes based on visible/near-infrared spectra,” Anal. Chim. Acta, vol. 584, no. 2, pp. 379–384, 2007.

Leave a Reply

Your email address will not be published. Required fields are marked *