A Miniaturized and Low-Cost Near-Infrared Spectroscopy Measurement System for Alfalfa Quality Control

Round 1

Reviewer 1 Report

The paper is good and had interesting results.
I miss in the two firsts paragraph I don’t see one reference about the Ideas showed, I think can improve it. Or other paragraphs in the introduction can cited same reference.

Author Response

Reviewer #1

The paper is good and had interesting results. I miss in the two firsts paragraph I don’t see one reference about the Ideas showed, I think can improve it. Or other paragraphs in the introduction can cite same reference.

Answers to Reviewer 1

Thank you for this comment. Following the Reviewer´s comment, four references have been included in the two first paragraphs.

Reviewer 2 Report

The reviewed article, judging by the title, should present a Miniaturized Optoelectronic Measurement System for Decentralized Agrifood Quality Control. However, the title does not reflect the real content and the Manuscript presents only the results of measurement and modeling of alfalfa hay, performed on ready-made equipment.

In the Introduction section, it is necessary to use more modern publications, since, for example, references [1-6] are 15-20 years old, and [7-9] are 15-30 years old. Of the 23 references in the Manuscript, only 5 are under five years old.

Currently, there are quite a large number of NIR sensors of products, including feed. In the Introduction section, the authors do not provide their comparative analysis, including in terms of size and cost. There is no justification for the development of the device, which casts doubt on the expediency of the authors' work.

How were the moisture content of the samples, the possible presence of impurities, the degree of homogeneity, and the degree of grinding taken into account in experimental studies? Nothing is said about this in the Manuscript, although it is likely to influence the results.

How to analyze heterogeneous feeds (silage, compound feed and other feed mixtures) with a gap size (10*10mm)? How and by how much can the size of the gap be increased? How does the illumination of the samples change? How does the noise level change?

Despite the stated price (line 50), the cost of the "Measurement System" was not calculated anywhere.

The conclusions (lines 321-325) are not supported by the materials of the Manuscript. There is not a single numerical value in the Abstract and Conclusion confirming the authors' statements.

 

Editorial comments:

1. Lines 31-32. This phrase should be deleted.

2. It is better to combine Sections 2 and 3, since "3.1 Forage Samples" are not methods, but materials.

3. The phrase (lines 313-314) does not apply to Conclusions. This is rather an Abstract.

 

Unfortunately, in the actual materials of the Manuscript, the "Measurement System", capable of competing with existing devices, is not presented. In this form, I consider the Manuscript unsuitable for publication.

Author Response

Reviewer 2

Thank you for your highly positive comments, which will contribute to improving the quality of the manuscript. All comments were carefully considered, and the manuscript was revised accordingly.

Comment #1: The reviewed article, judging by the title, should present a Miniaturized Optoelectronic Measurement System for Decentralized Agrifood Quality Control. However, the title does not reflect the real content and the Manuscript presents only the results of measurement and modeling of alfalfa hay, performed on ready-made equipment.

The reviewer was right. In order to better reflect the content of the manuscript we have renamed the paper “A Miniaturized and Low-Cost NIRS Measurement System for Alfalfa Quality Control”.

Comment #2: In the Introduction section, it is necessary to use more modern publications, since, for example, references [1-6] are 15-20 years old, and [7-9] are 15-30 years old. Of the 23 references in the Manuscript, only 5 are under five years old.

Thank you for your comment. We agree, references need to be updated. Based on your suggestion, we have updated references 1-6 and 8, 9. Reference 7 is the “Handbook of Near Infrared Spectroscopy third edition”, which contains reference data to interpret spectra and nutritive values. At this time, the updated fourth edition (2021) is not disposable in our research group.

Comment #3: Currently, there are quite a large number of NIR sensors of products, including feed. In the Introduction section, the authors do not provide their comparative analysis, including in terms of size and cost. There is no justification for the development of the device, which casts doubt on the expediency of the authors' work.

Thank you for this comment. Following this comment, in the revised article a table (Table 1) was included to compare the main specifications of some commercial miniaturized spectrometers, as found on the web sites of the manufacturers.

Comment #4: How was the moisture content of the samples, the possible presence of impurities, the degree of homogeneity, and the degree of grinding taken into account in experimental studies? Nothing is said about this in the Manuscript, although it is likely to influence the results.

Thank you for your comment. We agree that all those parameters: impurities, homogeneity, and grinding are key factors when developing a NIR assay. It is important to include all the disposable variability of the samples, including those factors mentioned before (homogeneity, grinding) and quality (this parameter is linked to impurities) to minimize outliers. Laboratories have grinding systems, however to develop a low-cost system, we have used a conventional seed mill that is cheap and easy to obtain by farmers or technicians. The particle size after milling is variable, and this variability has been included in the spectra and in those developed chemometric models. In order to clarify this information we have included a new paragraph in the text (Section 2.3 Forage Samples):

“In this study, a calibration set of 57 samples of hay or dehydrated alfalfa collected in the north of Spain was involved. Approximately 200g of alfalfa were collected during sampling procedure to be analyzed using miniaturized NIRs. Prior to collecting raw scans, samples were homogenized. After that, alfalfa samples were milled using a domestic spice mill (cheap and easy to use) and re-scanned in their ground form. This type of mill does not allow mesh size setting. The variability associated with this factor appears in collected spectra, and also in chemometric results. This is because all spectroscopic information is considered when developing calibrations.”

“After finishing NIR analysis, the quality of alfalfa samples, based on animal feed requirements, was characterized according to their nutritive value parameters. This was done using laboratory reference procedures. A Van Soest analysis [16] was performed on neutral detergent fiber (NDF), a gravimetric analysis was conducted on mineral content (MC), and a Kjeldahl analysis was conducted on crude protein (CP). CP and NDF are two of the critical parameters to classify alfalfa quality. CP values of 19% or higher, combined with crude fiber values of 29 % or lower are related to good forage quality [17]. Table 2 summarizes the statistics for nutritive parameters of all the samples involved in this study. It includes the range and variability of each analyzed parameter. Annex I includes individual values for each alfalfa sample involved in this study.”

Comment #5: How to analyze heterogeneous feeds (silage, compound feed and other feed mixtures) with a gap size (10*10mm)? How and by how much can the size of the gap be increased? How does the illumination of the samples change? How does the noise level change?

Thank you for your comment, the design of the Miniaturized NIRS included in this study includes a sample cup of 10´10 mm. WE agree with you, it is possible the design other gap sizes, or optical fibers to increase scanning window. With a bigger scanning window you improve the scanning surface, obtaining more sample information in only one scan. However, in order to minimize the effect of the gap size, after homogenization, three subsamples and ten spectra per subsample were scanned on the Miniaturized NIRs. The total number of scans collected per sample was 3*10=30 scans.

About the illumination effect, this topic has not been studied in this work, because we have employed only the designed sample cup, but, when increasing scanning window, the number of subsamples to be collected is reduced, and the analysis time too. And depending of the type of the designed cap the gap size can be find wide variability.

In order to explain this topic and clarify results, in Material and Methods Section (2.4 Spectral Acquisition) we have included the following text:

“In order to increase the sampling window and improve the spectroscopic information for each sample, after homogenizing alfalfa samples, each was divided into three subsamples and scanned on miniaturized NIRs. Each spectrum is an average of 10 spectra in a wavelength range between 901-1700 nm, with a non-linear path wavelength between 2.9 and 3.9 nm. A total of 30 scans were collected for each alfalfa sample.”

Comment #6: Despite the stated price (line 50), the cost of the "Measurement System" was not calculated anywhere.

The reviewer is right. In Section 2 of the revised article, the price of the main components is included:

• NIRscan Nano Evaluation Module: $999 (website)
• Microcontroller (LOLIN ESP32): $10 (website)
• Servomotor (MG90S): $5
• DC-DC converter (Pololu U1V0F5): $6

Comment #7: The conclusions (lines 321-325) are not supported by the materials of the Manuscript. There is not a single numerical value in the Abstract and Conclusion confirming the authors' statements.

You're correct. I appreciate your suggestion. Conclusion Section was rewrite. Lines 321 to 324 have been modified as follows:

“The proposed instrumentation allows users to evaluate the quality of the forage, increase sampling without incurring a cost, obtain results in real time, and make quick decisions, avoiding delays related to carrying samples from the farm to the laboratory. Moreover, specialized training is not required for users of this instrument”.

Reviewer 3 Report

 

The paper presents a low-cost and easy-to-use optoelectronic measurement system based on near-infrared spectrometry (NIRS) technology to improve feed chain quality and safety. The system utilizes the Texas Instruments´ NIRscan Nano Evaluation Module (EVM) spectrometer, which has a large sensing area and high resolution suitable for forage samples. The objective of the research is not only to check the efficiency of the proposed system but also to determine its characteristics and explore ways to improve it. Forage quality and safety are critical in animal husbandry, and the paper focuses on analyzing key quality parameters like fiber content, mineral content, and protein content using NIRS techniques.

1.      The paper addresses a significant challenge faced by dairy farm technicians and producers, who require rapid and reliable knowledge of forage quality on their farms. It highlights the importance of food safety and quality in the initial stages of the food chain.

2.      The use of the Texas Instruments´ NIRscan Nano Evaluation Module (EVM) spectrometer as the core of the measurement system offers a low-cost solution, making it more accessible for dairy farm technicians and producers.

3.      The development of a miniaturized NIR system, which is easy to use and specialized in quality control of raw forages, is an important advancement. This makes it feasible for on-site and real-time analysis.

4.      The research explores the effect of sampling presentation (raw or ground) on calibration statistics. Even though homogeneous forage samples (ground) provide better calibration models, satisfactory results can still be obtained for all sampling procedures.

5.      The paper emphasizes that the proposed system's characteristics are in line with precision farming requirements, as it allows quick on-site analysis by non-specialized personnel and provides real-time responses.

6.      The authors suggest the possibility of future Internet access for sharing information between different NIRscan nano instruments. Standardization and sharing of calibration models can lead to better sample variability and improved calibration statistics.

Author Response

Reviewer 3

The paper presents a low-cost and easy-to-use optoelectronic measurement system based on near-infrared spectrometry (NIRS) technology to improve feed chain quality and safety. The system utilizes the Texas Instruments´ NIRscan Nano Evaluation Module (EVM) spectrometer, which has a large sensing area and high resolution suitable for forage samples. The objective of the research is not only to check the efficiency of the proposed system but also to determine its characteristics and explore ways to improve it. Forage quality and safety are critical in animal husbandry, and the paper focuses on analyzing key quality parameters like fiber content, mineral content, and protein content using NIRS techniques.

1. The paper addresses a significant challenge faced by dairy farm technicians and producers, who require rapid and reliable knowledge of forage quality on their farms. It highlights the importance of food safety and quality in the initial stages of the food chain.
2. The use of the Texas Instruments´ NIRscan Nano Evaluation Module (EVM) spectrometer as the core of the measurement system offers a low-cost solution, making it more accessible for dairy farm technicians and producers.
3. The development of a miniaturized NIR system, which is easy to use and specialized in quality control of raw forages, is an important advancement. This makes it feasible for on-site and real-time analysis.
4. The research explores the effect of sampling presentation (raw or ground) on calibration statistics. Even though homogeneous forage samples (ground) provide better calibration models, satisfactory results can still be obtained for all sampling procedures.
5. The paper emphasizes that the proposed system's characteristics are in line with precision farming requirements, as it allows quick on-site analysis by non-specialized personnel and provides real-time responses.
6. The authors suggest the possibility of future Internet access for sharing information between different NIRscan nano instruments. Standardization and sharing of calibration models can lead to better sample variability and improved calibration statistics.

We appreciate very much the highly positive comment from this Reviewer. The Reviewer did not place any question/remark and recommended the manuscript to be accepted as submitted without revision.

Reviewer 4 Report

The paper Miniaturized Optoelectronic Measurement System for Decentralized Agrifood Quality Control by Candela Melendreras 1 , Ana Soldado 1,*, José M. Costa-Fernández 1 , Alberto López 2 and Francisco Ferrero 2,* 4 presents an interesting new tool for feed quality control. As a proposal, it is well documented and scientifically sound. One observation would be: this method needs a library data that should contain the main characteristics  required in order to qualify the feed as acceptable for delivery to animals.

 

The English should be revised, To grind is an irregular verb and its forms are tricky. Some other formulations should be carefully read and corrected.

Author Response

Reviewer 4

The paper Miniaturized Optoelectronic Measurement System for Decentralized Agrifood Quality Control by Candela Melendreras 1, Ana Soldado 1,*, José M. Costa-Fernández 1 , Alberto López 2 and Francisco Ferrero 2,* 4 presents an interesting new tool for feed quality control.

As a proposal, it is well documented and scientifically sound. One observation would be: this method needs a library data that should contain the main characteristics required in order to qualify the feed as acceptable for delivery to animals.

Thank you for your comment. We agree with you, and according to your suggestion, we have included the entire library including reference data for each sample and statistics (average, maximum, minimum and standard deviation) in Annex I of the reviewed version.

In response to your suggestion, we have added an additional reference (Reference 17, Lacefield, Garry D., "Alfalfa Hay Quality Makes the Difference". Agriculture and Natural Resources. 1988 Publications. 32. https://uknowledge.uky.edu/anr_reports/32), and the following paragraph in subsection 2.2:

“After finishing NIR analysis, the quality of alfalfa samples, based on animal feed requirements, was characterized according to their nutritive value parameters. This was done using laboratory reference procedures. A Van Soest analysis [16] was performed on neutral detergent fiber (NDF), a gravimetric analysis was conducted on mineral content (MC), and a Kjeldahl analysis was conducted on crude protein (CP). CP and NDF are two of the critical parameters to classify alfalfa quality. CP values of 19% or higher, combined with crude fiber values of 29 % or lower are related to good forage quality [17]. Table 2 summarizes the statistics for nutritive parameters of all the samples involved in this study. It includes the range and variability of each analyzed parameter. Annex I includes individual values for each alfalfa sample involved in this study.”

Round 2

Reviewer 2 Report

The authors took into account the Reviewer's comments and significantly improved the quality of the Manuscript. For a better perception by readers, I recommend:

1. Citations [5-8] and [9-12] should be divided into several citations by 1-3 links.

2. Improve the quality of Figures 4, 5.

Author Response

Thank you for your highly positive suggestions, which will contribute to improving the quality of the manuscript. All suggestions were carefully considered, and the manuscript was revised accordingly.

The authors took into account the Reviewer's comments and significantly improved the quality of the Manuscript. For a better perception by readers, I recommend:

1. Citations [5-8] and [9-12] should be divided into several citations by 1-3 links.

Thank you for this suggestion. Following the Reviewer´s comment, citations [5-8] and [9-12] were shown individually. Two of them were deleted.

2. Improve the quality of Figures 4, 5.

You are right. Thank you for this suggestion. The quality of Figures 4, 5 have been improved.