Corrosion factor detection system, corrosion factor detection method, and program

Abstract

To specify an attribute of a microorganism group affecting corrosion of a metal material.SOLUTION: The corrosion factor detection system measures a sample S including a test liquid L containing microorganisms and a metal material M in contact with the test liquid L to derive a corrosion process reflecting physical quantity which is a physical quantity whose value changes according to the corrosion process of the metal material M. In addition, the corrosion factor detection system derives first microorganism information which is information capable of specifying at least one of the quantity and quality of microorganisms contained in the test liquid L. The corrosion factor detection system then derives attribute information on microorganisms that can affect corrosion of the metal material M based on the time-series data on the corrosion process reflection physical quantity and the time-series data on the first microorganism information.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present disclosure relates to a corrosion factor detection system, a corrosion factor detection method, and a program. [Background technology]

[0002] Metal materials can be corroded by microorganisms. Patent Documents 1 and 2 describe techniques for detecting microbial corrosion of metal materials.

[0003] Patent Document 1 discloses a method for detecting microbial corrosion in a metal pipe through which a liquid is flowing. Specifically, Patent Document 1 discloses a method for determining whether a biofilm containing microorganisms that cause microbial corrosion has been formed in the metal pipe based on the results of comparing the change per unit time in the content of microorganisms, which are substances present in the liquid flowing through the metal pipe and are derived from bacteriophages that infect the microorganisms, with a reference value.

[0004] Patent Document 2 discloses that in a system in which a metal comes into contact with water, changes in the potential of the metal due to the adhesion of microbial-based fouling are measured over time, and changes in the oxidation of the water are measured over time, and based on the results of these measurements, it is determined whether there is an increased possibility of corrosion of the metal material and whether the cause of the corrosion is due to the adhesion of microbial-based fouling or an increase in oxidation in the system. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] Patent No. 7267588 [Patent Document 2] Patent No. 3862122 [Non-patent literature]

[0006] [Non-Patent Document 1] Gavin M. Douglas, Vincent J. Maffei, Jesse R. Zaneveld, Svetlana N. Yurgel, James R. Brown, Christopher M. Taylor, Curtis Huttenhower, Morgan GI Langille"PICRUSt2 for prediction of metagenome functions" Nature Biotechnology volume 38, pages685-688 (2020) [Non-patent document 2] Kanehisa Laboratories, "KO (KEGG ORTHOLOGY) Database," [Retrieved December 17, 2024], Internet<https: / / www.genome.jp / kegg / ko.html> Summary of the Invention [Problem to be solved by the invention]

[0007] However, the techniques described in Patent Documents 1 and 2 cannot identify the attributes of the microbial community that affect the corrosion of metal materials.

[0008] The present disclosure has been made in consideration of such problems, and aims to make it possible to identify the attributes of microbial communities that affect the corrosion of metal materials. [Means for solving the problem]

[0009] The corrosion factor detection system disclosed herein comprises a first derivation means for deriving a corrosion process-reflecting physical quantity, which is a physical quantity whose value changes depending on the corrosion process of the metal material, by performing measurements on a sample including a medium containing a microbial group and a metal material in contact with the medium; a second derivation means for deriving first microbial group information, which is information that can identify at least one of the quantity and quality of the microbial group contained in the medium; and a third derivation means for deriving attribute information of the microbial group that may affect the corrosion of the metal material based on time series data of the corrosion process-reflecting physical quantity derived by the first derivation means and time series data of the first microbial group information derived by the second derivation means.

[0010] The corrosion factor detection method disclosed herein comprises a first derivation step of deriving a corrosion process-reflecting physical quantity, which is a physical quantity whose value changes depending on the corrosion process of the metal material, by performing measurements on a sample including a medium containing a microbial group and a metal material in contact with the medium; a second derivation step of deriving first microbial group information, which is information that can identify at least one of the quantity and quality of the microbial group contained in the medium; and a third derivation step of deriving attribute information of the microbial group that may affect the corrosion of the metal material based on time series data of the corrosion process-reflecting physical quantity derived by the first derivation step and time series data of the first microbial group information derived by the second derivation step.

[0011] The program of the present disclosure causes a computer to function as each of the means of the corrosion factor detection system. [Effects of the Invention]

[0012] According to the present disclosure, it is possible to identify the attributes of microorganisms that affect the corrosion of metal materials. [Brief explanation of the drawings]

[0013] [Figure 1] FIG. 1 is a diagram illustrating an example of the configuration of a corrosion factor detection system. [Figure 2] FIG. 1 illustrates an example of a test device. [Figure 3]FIG. 1 is a diagram showing an example of a culture medium configuration in tabular form. [Figure 4] FIG. 1 is a diagram showing an example of time-series data of the natural potential of a metal material. [Figure 5] FIG. 2 is a diagram illustrating an example of a functional configuration of an information processing device. [Figure 6A] 1 is a flowchart showing an example of a corrosion factor detection method (a part where a preliminary test is performed). [Figure 6B] 1 is a flowchart showing an example of a corrosion factor detection method (a part that performs prediction). [Figure 7] FIG. 1 shows an example of the change over time in the relative proportion of each OTU. [Figure 8] FIG. 1 is a diagram showing a first example of measured values, total estimated values, and CV estimated values ​​of the natural potential of a metallic material. [Figure 9] FIG. 10 is a diagram showing a second example of the measured value, total estimated value, and CV estimated value of the natural potential of a metallic material. DETAILED DESCRIPTION OF THE INVENTION

[0014] Hereinafter, embodiments of the present disclosure will be described with reference to the drawings. The term "same length, position, size, spacing, etc.," as used herein, refers not only to those that are exactly the same, but also to those that are different within the scope of the present disclosure (for example, those that are different within the tolerances determined at the time of design). For ease of explanation, only the portions necessary for explanation are shown in simplified form in each drawing as necessary.

[0015] [First embodiment] First, the first embodiment will be described. FIG. 1 is a diagram showing an example of the configuration of a corrosion factor detection system according to this embodiment. FIG. 2 is a diagram showing an example of a testing device 200. The corrosion factor detection system is an example of a system for deriving a microbial community that may affect the corrosion of a metal material M. Here, "a microbial community that may affect the corrosion of a metal material M" includes cases where it is clear that the microbial community will affect the corrosion of the metal material M (for example, it is described in literature that the microbial community will affect the corrosion of the metal material M). Furthermore, "a microbial community that may affect the corrosion of the metal material M" preferably includes cases where there is a possibility that the microbial community will affect the corrosion of the metal material M (for example, when a microbial community may affect the corrosion of the metal material M, it is unclear whether the microbial community will affect the corrosion of the metal material M). In FIG. 1, the present embodiment illustrates an example in which the corrosion factor detection system includes a DNA extraction device 110, a sequencer 120, a potential measurement device 130, and an information processing device 140.

[0016] In FIG. 1, this embodiment illustrates a case where a test liquid L is collected from a test device 200 illustrated in FIG. Here, an example of a testing apparatus 200 will be described. In FIG. 2, the testing apparatus 200 is an example of an apparatus for measuring a physical quantity reflecting a corrosion process. The physical quantity reflecting a corrosion process is a physical quantity whose value changes depending on the corrosion process of a metallic material. In this embodiment, a case is illustrated in which the physical quantity reflecting a corrosion process is derived by performing measurements on a sample including a medium containing a microorganism group and a metallic material in contact with the medium. The corrosion process of the metallic material M includes a stage before the corrosion of the metallic material M is clearly apparent. During the corrosion process of the metallic material M (until the corrosion is clearly apparent), changes in the surface morphology of the metallic material M occur as the process progresses. FIG. 2 illustrates a case in which the liquid medium is a test liquid L, the metallic material is a metallic material M, and the sample is a sample S (specific examples of the test liquid L, the metallic material M, and the sample S will be described later). The physical quantity reflecting a corrosion process may be a physical quantity related to the quality of the metallic material M, a physical quantity related to the quantity of the metallic material M, or both. In this embodiment, the corrosion process-reflecting physical quantity is the natural potential of a metal material M. However, the corrosion process-reflecting physical quantity is not limited to the natural potential of a metal material M (this will be described later). In the following description, the natural potential of a metal material will be abbreviated to natural potential as necessary.

[0017] FIG. 2 illustrates an example in which the testing apparatus 200 includes lead wires 210 and 260, a tube 220, a reference electrode 230, a container 240, and a stopper 250. The sample S includes a test liquid L and a metal material M in (direct) contact with the test liquid L. It is sufficient that at least a portion of the metal material M is in contact with the test liquid L. However, it is preferable that as large an area of ​​the metal material M as possible be in contact with the test liquid L. This is because the corrosion process of the metal material M is more likely to be clearly reflected by the value of the physical quantity reflecting the corrosion process (the natural potential in this embodiment). For example, it is preferable that 40% or more of the surface area of ​​the metal material M be in contact with the test liquid L, more preferably 60% or more, and even more preferably 80% or more. FIG. 2 illustrates an example in which the metal material M and the test liquid L are contained in a container 240.

[0018] The type of metal material M is not limited, but may be, for example, stainless steel, low alloy steel, nickel, nickel alloy, titanium, titanium alloy, copper, copper alloy, chromium, chromium alloy, molybdenum, molybdenum alloy, tungsten, tungsten alloy, etc.

[0019] FIG. 2 illustrates an example in which one end of the lead wire 210 is electrically connected to the metal material M and the other end of the lead wire 210 is connected to the first measurement terminal of the potential measuring device 130. One end of the lead wire 210 is electrically connected to the metal material M by soldering or the like. It is preferable that the connection portion between the metal material M and the lead wire 210 does not directly contact the test solution L. For example, the area including the connection portion between the metal material M and the lead wire 210 may be coated with an insulating resin such as epoxy resin. It is also preferable that the conductor portion of the lead wire 210 does not directly contact the test solution L. As illustrated in FIG. 2, the lead wire 210 may be placed inside a non-metallic tube 220. Note that, for example, if the lead wire 210 is coated with an insulating coating, the lead wire 210 does not need to be placed inside the tube 220 described above. In this case, the test device 200 does not need to include the tube 220.

[0020] 2 illustrates an example in which a reference electrode 230 is placed in a container 240 in contact with a test solution L in order to measure the natural potential of a metal material M. The reference electrode 230 may be, for example, a double junction reference electrode (more specifically, for example, an Ag / AgCl double junction reference electrode) or another reference electrode.

[0021] FIG. 2 illustrates an example in which one end of the lead wire 260 is electrically connected to the reference electrode 230, and the other end of the lead wire 260 is connected to the second measurement terminal of the potential measuring device . 2 illustrates an example in which the opening of the container 240 is blocked with a plug 250. In this case, the tube 220 and the reference electrode 230 are placed inside the container 240, for example, while passing through a hole formed in the plug 250. It is preferable to limit the flow of oxygen between the inside and outside of the container 240 as much as possible.

[0022] The test liquid L is an example of a medium containing a microbial population. The sample liquid L is prepared, for example, by mixing soil containing a microbial population with a culture medium (e.g., a liquid culture medium). The test liquid L is a so-called microbial suspension. In this way, a solid such as soil may be contained in the medium. Here, the microbial population refers to a microbial population classified based on a certain trait or genotype. The genotype also includes the base sequence of the microbial genome. For example, microorganisms classified into the same group by genus, species, subspecies, etc. may be considered to be the same microbial population, but the classification method of the microbial population is not limited to these. In this embodiment, a case where the microbial populations classified into the same group are a single type of microbial population is exemplified. In other words, a case where multiple microbial populations classified into different groups are different types of microbial populations is exemplified.

[0023] The soil contained in test liquid L may be soil from any location. For example, if there is a location known to be prone to corrosion of metal materials, the soil contained in test liquid L may be soil from that location. The soil contained in test liquid L may be soil from a single location or a mixture of soils from multiple locations. Furthermore, if a metal material of the same type as metal material M has already been installed, the soil contained in test liquid L may be soil at the location where the metal material will be installed. Furthermore, for example, if there is a location where an object having the same type of metal material as metal material M is planned to be installed, the soil contained in test liquid L may be soil from that location (the planned installation location). Even in these cases, the soil contained in test liquid L may be soil from a location different from the installed location and planned installation location of the metal material of the same type as metal material M. This is because, for example, microbial communities that may affect the corrosion of the metal material may become present in the installed and planned installation locations of the metal material in the future.

[0024] The culture medium contained in the test liquid L is used for the purpose of growing (preferably multiplying) the microbial community. In order to reduce the effects of corrosion of the metal material M due to factors other than the microbial community, the amount of chlorine and sulfate ions contained in the culture medium is preferably as small as possible, and most preferably 0 (zero). Figure 3 shows an example of the composition of the culture medium. In the test liquid L, the volume ratio of soil may be smaller, larger, or the same as the volume ratio of the culture medium. For example, the volume ratio of soil:volume ratio of culture medium may be 1:3. Furthermore, for example, if the soil contains nutrients necessary for the survival (preferably multiplication) of the microbial community, the culture medium may not be contained in the test liquid L.

[0025] Furthermore, multiple test devices 200 may be fabricated. In this case, the fabrication conditions for the multiple test devices 200 are preferably similar, and most preferably identical. For example, multiple metallic materials M may be collected from the same base material. Alternatively, the test liquid L may be divided and placed in multiple containers 240. In this manner, for example, the spontaneous potential of the metallic material M can be measured in parallel under approximately the same conditions (preferably the same conditions) in each of the multiple test devices 200. In this case, the timing at which the multiple test devices 200 start testing (the timing at which the metallic material M is immersed in the test liquid L) is preferably similar, and more preferably identical. Furthermore, the measured values ​​of the spontaneous potential of the metallic material M in the multiple test devices 200 may be expressed as relative values ​​from an initial value. In this way, the measured values ​​of the spontaneous potential of the metallic material M in the multiple test devices 200 can be compared without performing a process to unify the reference values ​​of the measured values ​​of the spontaneous potential of the metallic material M in the multiple test devices 200.

[0026] In order to further reduce the influence of corrosion of the metal material M due to factors other than the microbial population, for example, the soil may be filtered to collect only the microbial population in the soil, and the test solution L may be prepared by mixing the collected microbial population with the above-mentioned culture medium. Furthermore, for example, the following steps may be repeated for a certain period of time: bringing a metal material M into contact with (immersing in) soil from which microorganisms have been filtered (i.e., soil from which microorganisms have been removed), removing the metal material M, checking whether corrosion has occurred on the metal material M, and, if corrosion has not occurred on the metal material M, bringing the metal material M into contact with the soil from which microorganisms have been filtered again. If no corrosion occurs on the metal material in contact with the soil from which microorganisms have been filtered, it can be determined that the metal material M is not corroded by factors other than the microorganisms. On the other hand, if corrosion occurs on the metal material in contact with the soil from which microorganisms have been filtered, it can be determined that factors other than the microorganisms are involved in the corrosion of the metal material M.

[0027] Note that if the cause of corrosion of the metal material M is a factor other than microorganisms (e.g., chlorine), it is considered that the information processing device 140, described below, will not derive microorganisms that may affect the corrosion of the metal material M. Therefore, when measuring the natural potential of the metal material M, measures to reduce factors other than microorganisms as factors causing the corrosion of the metal material M do not need to be taken. However, for example, if the information processing device 140 cannot eliminate the possibility that a microorganism having a spurious correlation with the above-mentioned physical quantity reflecting the corrosion process (natural potential in this embodiment) may be derived as a microorganism that may affect the corrosion of the metal material M, measures to reduce factors other than microorganisms as factors causing the corrosion of the metal material M may be taken. Note that in the above example, measures to reduce factors other than microorganisms as factors causing the corrosion of the metal material M include not including chlorine-containing substances in the internal solution and culture medium of the reference electrode 230 and filtering the soil.

[0028] Returning to the explanation of FIG. 1, the DNA extraction apparatus 110 is an apparatus for extracting DNA fragments of a microorganism group. For example, the DNA extraction apparatus 110 may collect a portion of the test solution L in the container 240, centrifuge the collected test solution L to recover a precipitate in the test solution L, and extract the DNA fragments contained in the precipitate using a DNA extraction kit. The results illustrated in FIG. 7, which will be described later, were obtained by extracting DNA fragments of a microorganism group contained in the test solution L using "ISOIL for Beads Beating" manufactured by Nippon Gene Co., Ltd.

[0029] When sampling a portion of the test liquid L in the container 240, for example, the stopper 250 may be (temporarily) removed from the container 240, and a portion of the test liquid L in the container 240 may be sampled. Alternatively, a stopper (not shown) may be provided on the side or bottom of the container 240, and the stopper may be (temporarily) removed from the container 240, and a portion of the test liquid L in the container 240 may be sampled.

[0030] Furthermore, to further suppress fluctuations in the spontaneous potential due to opening the container 240, a container for collecting the test liquid L may be prepared separately from the test device 200. In this case, first, the test liquid L is divided into the container 240 of the test device 200 and a container for collecting the test liquid L. Then, two metal materials M cut from the same base material are placed in the container 240 of the test device 200 and the container for collecting the test liquid L, one each, and brought into contact with the test liquid L. The size and shape of the metal material M to be placed in the container 240 of the test device 200 and the size and shape of the metal material M to be placed in the container for collecting the test liquid L are preferably similar, more preferably approximately the same, and most preferably the same. Furthermore, the timing at which the metal material M is brought into contact with the test liquid L in the container 240 of the test device 200 and the timing at which the metal material M is brought into contact with the test liquid L in the container for collecting the test liquid L are preferably similar, more preferably approximately the same, and most preferably the same. If a container for collecting the test liquid L is prepared separately from the test device 200, it is possible to collect the test liquid L having conditions similar to those of the test liquid L in the container 240 of the test device 200 at any time without opening the container 240 of the test device 200. The opening of the container for collecting the test liquid L may be blocked with a stopper or the like.

[0031] The sequencer 120 is a device that determines the base sequence of a DNA fragment. The sequencer 120 may be equipped with a next-generation sequencer. The base sequence of the microbial community contained in the test solution L may be derived by performing analysis using the sequencer 120. In this case, the DNA fragment may be purified and its concentration may be measured, and PCR amplification may be performed using primers. In this case, the sequencer 120 may derive data on the base sequence of the microbial community contained in the test solution L by performing analysis using the PCR product obtained in this manner. When attempting to phylogenetically classify microorganisms, the target may be, for example, the ribosomal RNA gene (rRNA gene or rDNA) of the microorganism, or other genes (e.g., cDNA). In metagenomic analysis, the entire gene may be decoded, or a part of it may be decoded. The results illustrated in FIG. 7, which will be described later, were obtained using "Veriti" (Veriti) by Thermo Fisher Scientific Co., Ltd. TM These results were obtained by deriving the base sequence data of the microbial group contained in test solution L using a 96-Well Thermal Cycler and an Ion GeneStudio S5 system.

[0032] Furthermore, as will be described later, for example, when a metal material of the same type as the metal material M has already been installed or when there is a location where a metal material of the same type as the metal material M is planned to be installed, the information processing device 140 may derive corrosion prediction information regarding the possibility of corrosion of the metal material of the same type as the metal material M, in addition to deriving the microbial community that may affect the corrosion of the metal material M. In this case, extraction of DNA fragments and derivation of base sequence data of the microbial community are performed in each case, when deriving the microbial community that may affect the corrosion of the metal material M and when deriving the corrosion prediction information. Note that the soil from which the microbial community is collected to derive the corrosion prediction information is the soil at the installation location or planned installation location of the metal material of the same type as the metal material M. On the other hand, the soil from which the microbial community is collected to derive the microbial community that may affect the corrosion of the metal material M may be the soil at the installation location or planned installation location of the metal material of the same type as the metal material M, or it may be soil at a location other than the installation location of the metal material.

[0033] The method for extracting DNA fragments from soil and deriving the base sequence data of the microbial community contained in the soil can be realized by known techniques. Therefore, the method for extracting DNA fragments from soil and deriving the base sequence data of the microbial community contained in the soil is not limited to the method described above.

[0034] The potential measuring device 130 is a device for measuring the natural potential of the metal material M in the testing device 200 (container 240). The natural potential of the metal material M is represented, for example, by the potential difference between the first measuring terminal and the second measuring terminal of the potential measuring device 130. As described above, this embodiment illustrates a case in which the first measuring terminal of the potential measuring device 130 is electrically connected to the metal material M, and the second measuring terminal of the potential measuring device 130 is electrically connected to the reference electrode 230. This embodiment illustrates a case in which the first derivation means is realized by using the potential measuring device 130.

[0035] In this embodiment, a case is illustrated in which the timing at which the test solution L is collected to derive the base sequence data of the microbial community and the timing at which the data on the natural potential of the metallic material M are obtained correspond to each other. FIG. 4 , which will be described later, illustrates a case in which the base sequence data of the microbial community and the data on the natural potential of the metallic material M are each obtained once a day. In this case, the data on the base sequence of the microbial community and the data on the natural potential of the metallic material M are obtained on the same day as data at the aforementioned corresponding times. For example, if the test solution L is collected on a certain day, the data on the natural potential of the metallic material M may be data measured on that day. In this case, the timing at which the natural potential of the metallic material M is measured is preferably as close as possible to the timing at which the test solution L is collected. Note that the timing at which the data on the base sequence of the microbial community and the data on the natural potential of the metallic material M are derived is not limited to once a day, but may be multiple times a day or once every several days. Furthermore, a function indicating the relationship between the natural potential of the metallic material M and time may be derived based on the data on the natural potential of the metallic material M. In this case, a value at any timing can be derived as the value of the natural potential of the metallic material M. Therefore, the timing for measuring the natural potential of the metallic material M does not have to coincide with the timing for collecting the test liquid L (in the example described above, these timings do not have to be on the same day).

[0036] As described above, a plurality of test devices 200 may be fabricated, and portions of the test liquid L may be collected sequentially from the plurality of test devices 200. In this way, shortages of the test liquid L may be prevented. Furthermore, shortages of the test liquid L may also be prevented by preparing a container for collecting the test liquid L separate from the test device 200 and collecting a portion of the test liquid L from the container. Note that, for example, if the size of the container 240 is sufficiently large and a sufficient amount of the test liquid L is present in the container 240, it is not necessary to fabricate a plurality of test devices 200 or to prepare a container for collecting the test liquid L, or both.

[0037] As described above, this embodiment illustrates a case where time-series data of the base sequences of a microbial community is derived using the sequencer 120. Also, this embodiment illustrates a case where time-series data of the natural potential of a metal material M is derived using the potential measuring device 130.

[0038] The information processing device 140 is an example of a device that performs processing to derive attribute information of a microorganism group that may affect the corrosion of the metal material M. In the following description, deriving attribute information of a microorganism group that may affect the corrosion of the metal material M will be referred to as a pre-test as necessary. Furthermore, this embodiment illustrates a case in which, after a pre-test, the information processing device 140 derives corrosion prediction information regarding the possibility of corrosion of a metal material of the same type as the metal material M that is the subject of the pre-test at an installation location or planned installation location of the metal material. In the following description, deriving the corrosion prediction information will be referred to as a prediction as necessary. Note that this embodiment illustrates a case in which the testing device 200 is not used during the prediction (when the corrosion prediction information is derived). Furthermore, for example, if there is no installation location or planned installation location for a metal material of the same type as the metal material M, the information processing device 140 may not need to perform a prediction (derive the corrosion prediction information).

[0039] The hardware of the information processing device 140 has, for example, one or more hardware processors such as a CPU (Central Processing Unit), one or more memories such as a RAM (Random Access Memory) and a ROM (Read Only Memory), and performs various operations by executing one or more programs stored in the memories by the one or more hardware processors. Furthermore, the hardware of the information processing device 140 has an input device and an output device.

[0040] Here, an example of the concept of this embodiment will be described. FIG. 4 is a diagram showing an example of time-series data of the natural potential of a metal material M. FIG. 4 illustrates a case where the metal material M is SUS304. Note that in FIG. 4, the natural potential values ​​are expressed as relative values ​​and are dimensionless quantities (the "(-)" in natural potential (-) indicates this). Also, in FIG. 4, natural potential data (values ​​derived by the potential measuring device 130) are indicated by black circles (●). Also, in FIG. 4, the day on which the number of elapsed days is 0 indicates the day on which the sample S was prepared (the day on which the metal material M was immersed in the test liquid L). In the following description, the day on which the number of elapsed days is 0 will be referred to as the test start date as necessary.

[0041] In Figure 4, it can be seen that the time series data of the natural potential of metal material M (SUS304) includes, in this order, a period in which the natural potential increases with the passage of time (a period from approximately 0 to 20 days), a period in which the natural potential remains approximately constant regardless of the passage of time (a period from approximately 20 to 50 days), and a period in which the natural potential decreases with the passage of time (a period from approximately 50 to 64 days).

[0042] For a certain period from the start of the test, the natural potential of the metallic material M generally increases over time, as illustrated in Figure 4 (see the period from approximately day 0 to day 20), regardless of the composition of the test solution L (the microbial community and the components of the test solution L). After that, the natural potential of the metallic material M shows a generally constant value regardless of the passage of time, which is thought to be due to a balance between the destruction and self-repair of the passive film on the surface of the metallic material M (see the period from approximately day 20 to day 50). After that, the natural potential of the metallic material M tends to decrease, which is thought to be due to an imbalance between the destruction and self-repair of the passive film on the surface of the metallic material M, and the amount of destruction of the non-conductive film begins to exceed the amount of self-repaired non-conductive film (see the period from approximately day 50 to day 64).

[0043] Furthermore, for the metal material M whose natural potential was measured as shown in FIG. 4, no clear corrosion was observed even after 64 days had passed since the start of the test. Genetic analysis (deriving base sequence data of the microbial community contained in the test solution L with which the metal material M was brought into contact) was performed for the period from the start of the test until 64 days had passed (see FIG. 7, which will be described later). Furthermore, the presence or absence of corrosion of the metal material M was visually confirmed on the 84th day from the start of the test, but no corrosion of the metal material M was confirmed. Therefore, when the surface of the metal material M was observed under a microscope, it was confirmed that small dark spots and small holes were present on the surface of the metal material M. These dark spots and holes are thought to be the starting points of corrosion (signs of corrosion).

[0044] From the above, it can be said that the natural potential of the metallic material M is a physical quantity whose value changes according to the corrosion process of the metallic material M. Therefore, the inventors of the present invention thought that the natural potential of the metallic material M can be used as a physical quantity that indicates signs of corrosion of the metallic material M. Such a physical quantity is not limited to the natural potential of the metallic material M, but may be any physical quantity whose value changes according to the corrosion process of the metallic material M. Therefore, as described above, such a physical quantity (physical quantity reflecting the corrosion process) is not limited to the natural potential of the metallic material M. For example, instead of or in addition to the natural potential of the metallic material M, the pH of the test solution L may be used as such a physical quantity (physical quantity reflecting the corrosion process).

[0045] When a microbial community is a factor in the corrosion of a metal material M, the value of a physical quantity (a physical quantity reflecting the corrosion process) whose value changes according to the corrosion process of the metal material M is thought to change according to at least one of the quantity and quality of the microbial community. Therefore, the physical quantity (a physical quantity reflecting the corrosion process) whose value changes according to the corrosion process of the metal material M is thought to change over time in response to time changes in information (first microbial community information) that can identify at least one of the quantity and quality of the microbial community. The inventors therefore considered that attribute information of a microbial community that may affect the corrosion of a metal material M could be derived based on time-series data of the physical quantity reflecting the corrosion process and time-series data of the first microbial community information. The first microbial community information is, for example, information including microbial species (types of microbial communities) derived from ribosomal RNA genes. The first microbial community information may be one type of information or multiple types of information. The first microbial community information may be a physical quantity or information other than a physical quantity. As information other than a physical quantity, the first microbial community information may be, for example, information that numerically represents the presence or absence of a microbial community. In this case, the first microbial community information may be information in which the value is 1 when a certain type of microbial community is present and the value is 0 when that type of microbial community is not present. The quality of the microbial community can be identified by identifying the type of microbial community from the value of such first microbial community information. Therefore, such first microbial community information is an example of information that can identify the quality of the microbial community. The first microbial community information may also be information that can identify the quantity of the microbial community. In this case, the quantity of the microbial community may be an absolute quantity (the quantity itself) or a relative quantity (relative ratio). The quantity of the microbial community may also be the quantity of each type of microbial community, or the quantity of one type of microbial community.

[0046] In addition, a method for deriving attribute information of a microbial group that may affect the corrosion of metal material M based on time series data of physical quantities reflecting the corrosion process and time series data of first microbial group information may be a rule-based method, etc.

[0047] Furthermore, in a machine learning model having a corrosion process-reflecting physical quantity as a response variable and first microbial community information as an explanatory variable, attribute information of a microbial community that may affect the corrosion of the metal material M may be derived by deriving (selecting) an explanatory variable (first microbial community information) that explains the response variable (corrosion process-reflecting physical quantity). For example, the type of microbial community corresponding to the explanatory variable (first microbial community information) selected based on evaluation indices such as the coefficient of determination and p-value for each combination of explanatory variables (first microbial community information) in the machine learning model (combination of multiple types of microbial communities) may be derived as attribute information of the microbial community that may affect the corrosion of the metal material M. In this case, the attribute information of the microbial community that may affect the corrosion of the metal material M is information that can identify the type of microbial community that may affect the corrosion of the metal material M. Note that the attribute information of the microbial community that may affect the corrosion of the metal material M is not limited to information that can identify the type of microbial community that may affect the corrosion of the metal material M, as long as it is information that can identify the attribute of the microbial community that may affect the corrosion of the metal material M.

[0048] The machine learning model may be, for example, a neural network. The machine learning model may also be, for example, a regression equation. In this case, attribute information of a microbial community that may affect the corrosion of a metal material may be derived based on the results of deriving regression coefficients of the regression equation based on time-series data of the corrosion process-reflecting physical quantities and time-series data of the first microbial community information. For example, attribute information of a microbial community that may affect the corrosion of a metal material may be derived using a variable selection method such as a stepwise method.

[0049] Here, the present inventors have proposed a method described in Japanese Patent No. 7299485. In the method described in Japanese Patent No. 7299485, first, a regression equation is used in which the rate of change in the amount of a specific substance is the target variable (dependent variable) and the content of a microbial group is the explanatory variable (independent variable), and important explanatory variables are selected based on the regression coefficients derived by performing a regression analysis that can reduce the regression coefficients to 0. Then, the microbial group (type) corresponding to the selected explanatory variable is identified as the microbial group (type) that can affect the rate of change in the amount of a specific substance.

[0050] The first microbial community information described above corresponds, for example, to the content of the microbial community in the method described in Japanese Patent No. 7299485. Furthermore, since the value changes due to the activity of the microbial community, the physical quantity reflecting the corrosion process can be said to correspond to the rate of change in the amount of a specific substance in the method described in Japanese Patent No. 7299485. Therefore, the present inventors believed that by applying the method described in Japanese Patent No. 7299485, using the physical quantity reflecting the corrosion process as the objective variable and the first microbial community information as the explanatory variable, it would be possible to accurately derive attribute information of the microbial community that can affect the corrosion of the metal material.

[0051] The corrosion factor detection system and corrosion factor detection method of this embodiment are based on the above idea. An example of an information processing device 140 included in the corrosion factor detection system of this embodiment and an example of a corrosion factor detection method using the information processing device 140 are described below. In the following, an example is given in which a corrosion process-reflecting physical quantity is used as a target variable and first microbial community information is used as an explanatory variable to derive attribute information of a microbial community that may affect the corrosion of a metal material, and the method described in Japanese Patent No. 7299485 is applied. However, the method for deriving attribute information of a microbial community that may affect the corrosion of a metal material may be any of the various methods exemplified in the description of the idea obtained by the present inventors, and is not limited to the method exemplified below.

[0052] FIG. 5 is a diagram illustrating an example of the functional configuration of the information processing device 140. FIG. 5 illustrates an example in which the information processing device 140 includes a preliminary test acquisition unit 511, a preliminary test microorganism deriving unit 512, a corrosion-causing microorganism deriving unit 513, a memory unit 521, a prediction acquisition unit 531, a prediction-time microorganism deriving unit 532, a corrosion prediction information deriving unit 533, and an output unit 541. The preliminary test acquisition unit 511, the preliminary test microorganism deriving unit 512, and the corrosion-causing microorganism deriving unit 513 are examples of parts for performing the above-mentioned preliminary test (deriving attribute information of a microorganism group that may affect the corrosion of the metal material M). The memory unit 521 is an example of a part for storing information indicating the results of the preliminary test. The prediction acquisition unit 531, the prediction-time microorganism deriving unit 532, and the corrosion prediction information deriving unit 533 are examples of parts for performing the above-mentioned prediction (deriving corrosion prediction information). The output unit 541 is an example of a unit for outputting information indicating the results of the prediction. FIG. 6A is a flowchart showing an example of a part of the corrosion factor detection method performed using the information processing device 140 that performs a preliminary test. FIG. 6B is a flowchart showing an example of a part of the corrosion factor detection method performed using the information processing device 140 that performs a prediction. An example of the functions of each unit will be described below. Note that the functions of the information processing device 140 may be realized by multiple devices. For example, a device that realizes at least one of the functions of the preliminary test acquisition unit 511, the preliminary test microorganism derivation unit 512, and the corrosion-causing microorganism derivation unit 513, the function of the memory unit 521, the functions of the prediction acquisition unit 531, the prediction microorganism derivation unit 532, and the corrosion prediction information derivation unit 533, and the function of the output unit 541 may be a device separate from the devices that realize the other functions.

[0053] (Pre-test acquisition unit 511, step S611) 6A, the pre-test acquisition unit 511 acquires time-series data of base sequences during the pre-test and time-series data of the free-space potential during the pre-test. In this embodiment, the time-series data of base sequences during the pre-test is time-series data of base sequences of a microorganism group obtained from the test solution L. In addition, in this embodiment, the time-series data of the free-space potential during the pre-test is time-series data of the free-space potential of the metal material M obtained by the potential measuring device 130.

[0054] For example, a transmitting / receiving device connected to the potential measuring device 130 may transmit time series data of the free-space potential of the metallic material M to the information processing device 140. In this case, the pre-test acquisition unit 511 may acquire the time series data of the free-space potential of the metallic material M by receiving the time series data of the free-space potential of the metallic material M transmitted from the transmitting / receiving device. In this case, communication between the transmitting / receiving device and the information processing device 140 may be wireless communication, wired communication, or communication via a network. Furthermore, for example, the time series data of the free-space potential of the metallic material M measured by the potential measuring device 130 may be stored in a storage medium. In this case, the pre-test acquisition unit 511 may acquire the time series data of the free-space potential of the metallic material M by reading the time series data of the free-space potential of the metallic material M from the storage medium.

[0055] Furthermore, for example, a transmission / reception device connected to the sequencer 120 may transmit time series data of the base sequences of the microorganisms to the information processing device 140. In this case, the preliminary test acquisition unit 511 may acquire the time series data of the base sequences of the microorganisms by receiving the time series data of the base sequences of the microorganisms transmitted from the transmission / reception device. In this case, communication between the transmission / reception device and the information processing device 140 may be wireless communication, wired communication, or communication via a network. Furthermore, for example, the time series data of the base sequences of the microorganisms derived by the sequencer 120 may be stored in a storage medium. In this case, the preliminary test acquisition unit 511 may acquire the time series data of the base sequences of the microorganisms by reading the time series data of the base sequences of the microorganisms from the storage medium.

[0056] (Pre-test microorganism extraction unit 512, step S612) 6A, the preliminary test microorganism derivation unit 512 derives time series data of the content of the microbial community at the time of the preliminary test as an example of time series data of the first microbial community information, based on the time series data of the base sequences of the microbial community at the time of the preliminary test. The time series data of the content of the microbial community is time series data for each type of microbial community.

[0057] In this embodiment, the preliminary test microorganism derivation unit 512 identifies the types of microorganisms contained in the test solution L based on the base sequences of the microorganisms obtained during the preliminary test and derives the content of each type of microorganism. Identifying the types of microorganisms contained in the test solution L based on the base sequences of the microorganisms can be achieved using known techniques. For example, as described in Japanese Patent No. 7299485, the types of microorganisms can be identified by performing a homology search against the Greengene 16S rRNA gene database for each representative OTU sequence and estimating the phylogenetic classification. It is not necessary to specifically identify the type of microorganism (e.g., the name of the microorganism) (this will be discussed later). Deriving the content of a microorganism based on the base sequences of the microorganisms can also be achieved using known techniques. The preliminary test microorganism derivation unit 512 may derive information corresponding to the content of the microorganisms, rather than the content of the microorganisms themselves.

[0058] For example, the preliminary testing microorganism derivation unit 512 may derive the relative content of each type of microorganism contained in the test solution L based on the base sequences of the microorganisms obtained during the preliminary testing. In this case, the preliminary testing microorganism derivation unit 512 may derive the number of microorganisms contained in the test solution L for each type of microorganism. In this case, the preliminary testing microorganism derivation unit 512 may derive the number of each type of microorganism contained in the test solution L for each type of microorganism based on the amount of DNA contained in the test solution L, or may derive the number of each type of microorganism contained in the test solution L for each type of microorganism based on the number of genes contained in the test solution L. The preliminary testing microorganism derivation unit 512 may derive the content of each type of microorganism by multiplying the relative content of each microorganism contained in the test solution L by the number of microorganisms contained in the test solution L. Furthermore, the preliminary testing microorganism derivation unit 512 may standardize the content of each type of microorganism thus derived and derive the content of each type of microorganism for each type of microorganism. Furthermore, the preliminary test microorganism derivation unit 512 may use the relative content ratio of the microorganisms contained in the test liquid L for each type of microorganism as the content of the microorganisms. In this way, the content of the microorganisms does not have to be the content of the microorganisms themselves, as long as it is information indicating the content of the microorganisms.

[0059] A specific example of a method for deriving the content of a microbial group is the method described in Japanese Patent No. 7299485. For example, nucleotide sequences may be analyzed using the QIIME (Quantitative Insights Into Microbial Ecology) pipeline. In this case, data quality and chimera are first checked, and for sequence data that meet predetermined criteria, highly similar sequence data (e.g., sequence data with 97% or more homology) are grouped into a single cluster. The most frequently occurring sequence among the sequences belonging to each cluster is designated as a representative OTU (Operational Taxonomic Unit) sequence. Using this representative OTU sequence, the presence and amount of each OTU may be treated as indicating the presence and amount of a single microbial group. Alternatively, the number of times each OTU is detected relative to the total number of OTUs detected may be used to derive the relative content of each microbial group. In this case, the number of genes of the microbial group (e.g., the number of eubacterial genes) contained in the test solution L is quantified using real-time PCR (e.g., QP-PCR (J-Bio21)). The amount of each OTU may then be derived as the amount of each microbial group contained in the test solution L by multiplying the relative proportion of each OTU by the number of genes in the microbial group contained in the test solution L. As described above, by grouping highly similar sequence data (e.g., sequence data with a homology of 97% or more) into one group, the group (cluster) is distinguished from other types of microbial groups as a group of the same type. Therefore, as described above, it is possible to identify each of multiple types of microbial groups from the representative OTU, etc., without deriving the specific type of microbial group (e.g., the name of the microbial group).

[0060] FIG. 7 is a diagram showing an example of the change over time in the relative proportion of each OTU derived as in the above specific example. In FIG. 7, day 0 indicates the test start date. Day 64 indicates the day 64 days have passed since the test start date. In the example shown in FIG. 7, 2795 types of microbial groups (OTUs) were detected. FIG. 7 shows an example of the relative proportions of the contents of these 2795 types of microbial groups. As described above, this embodiment illustrates a case in which the second derivation means is realized by using the preliminary test microorganism derivation unit 512. Furthermore, as described above, this embodiment illustrates a case in which the first microbial community information is information that can identify the quantity of a microbial community for each type of microbial community. In this case, the quality of a microbial community is identified by the type of microbial community, so the first microbial community information is information that can identify not only the quantity but also the quality of a microbial community.

[0061] (Corrosion-causing microorganism deriving unit 513, memory unit 521, steps S613 and S614) In step S613 of FIG. 6A, the corrosion-causing microorganism derivation unit 513 derives attribute information of a microorganism group that may affect the corrosion of the metal material M based on time-series data of the self-potential during the preliminary test and time-series data of the content of the microorganism group during the preliminary test. In the following description, the attribute information of a microorganism group that may affect the corrosion of the metal material M will be referred to as corrosion-causing microorganism attribute information as needed. Then, in step S614 of FIG. 6A, the storage unit 521 stores the corrosion-causing microorganism attribute information derived by the corrosion-causing microorganism derivation unit 513. The corrosion-causing microorganism attribute information may be any information that can identify the attributes of each microorganism group. The corrosion-causing microorganism attribute information may be, for example, the name or ID of the microorganism group. As described above, this embodiment illustrates a case in which the corrosion-causing microorganism derivation unit 513 derives corrosion-causing microorganism attribute information by using the method described in Japanese Patent No. 7299485. In the following description, a microorganism group that may affect the corrosion of the metal material M will be referred to as corrosion-causing microorganism group as needed.

[0062] The corrosion-causing microorganism derivation unit 513 extracts data for the same timing (the same day in this embodiment) from the time-series data of the free-space potential during the preliminary test and the time-series data of the content of the microbial community during the preliminary test, and creates data for one timing, for example, at multiple timings. In the example shown in FIGS. 4 and 7, the corrosion-causing microorganism derivation unit 513 extracts data for the same timing (the same day in this embodiment) from the time-series data of the free-space potential during the preliminary test and the time-series data of the content of the microbial community during the preliminary test, and uses the free-space potential during the preliminary test and the content of the microbial community during the preliminary test obtained on the same day as the data for that day, for each day from the test start date until 64 days after the test start date. In this way, the corrosion-causing microorganism derivation unit 513 creates a dataset including data for each timing (each day), for example.

[0063] The corrosion-causing microorganism derivation unit 513 may create a data set using all data obtained at the same time as the data on the natural potential during the preliminary test and the data on the content of the microbial population during the preliminary test, or may create a data set using only data from some of the data obtained at the same time.

[0064] Furthermore, the corrosion-causing microorganism deriving unit 513 may derive a function indicating the relationship between the free-space potential and time during the preliminary test based on time-series data of the free-space potential during the preliminary test. Similarly, the corrosion-causing microorganism deriving unit 513 may derive a function indicating the relationship between the content of the microbial community during the preliminary test and time based on time-series data of the content of the microbial community during the preliminary test. The corrosion-causing microorganism deriving unit 513 may derive the free-space potential and the content of the microbial community at the same timing in these functions as the free-space potential and the content of the microbial community at that timing.

[0065] 4, the spontaneous potential values ​​on the 4th and 5th days are extremely large. In this case, the corrosion-causing microorganism derivation unit 513 may treat the spontaneous potential on those days as an outlier and create a data set without using the data on those days.

[0066] As described above, for a certain period from the start of the test, the natural potential of the metal material M generally increases over time (see the period from approximately day 0 to day 20 in FIG. 4), regardless of the composition of the test solution L (the presence or absence of microbial communities and the components contained therein). Therefore, the corrosion-causing microorganism derivation unit 513 may create a data set without using data from such a period.

[0067] Furthermore, for example, if time-series data can be obtained at multiple timings as the time-series data of the spontaneous potential during the preliminary test and the time-series data of the content of the microbial community during the preliminary test, the derivation results of the corrosion-causing microbial community are unlikely to be significantly affected by data from a certain period after the start date of the test or by outlier data. Therefore, for example, in such a case, the corrosion-causing microorganism derivation unit 513 may create a dataset using all of the data obtained at the same timing as the spontaneous potential data during the preliminary test and the content of the microbial community during the preliminary test.

[0068] Using the data set obtained as described above, the corrosion-causing microorganism derivation unit 513 derives the type of corrosion-causing microorganism group as an example of corrosion-causing microorganism attribute information by the method described in Japanese Patent No. 7299485. In this embodiment, an example is shown in which the spontaneous potential of the metal material M is used as the objective variable (dependent variable) in place of the rate of change in the amount of a specific substance in the method described in Japanese Patent No. 7299485.

[0069] The method described in Japanese Patent No. 7299485 will be outlined below. First, the corrosion-causing microorganism derivation unit 513 creates a sample by resampling using the above-mentioned data set. Resampling methods include, for example, the Bootstrap method and the Jackknife method. The results shown in FIG. 8 (described later) were obtained by using the Bootstrap method. This type of resampling may also be performed when using the above-mentioned stepwise method.

[0070] Next, the corrosion-causing microorganism derivation unit 513 performs a penalized regression analysis on the sample created as described above, where the content of the microbial community is used as an explanatory variable (independent variable) and the spontaneous potential of the metal material M is used as a target variable (dependent variable), in which the regression coefficient can be reduced to 0. Examples of penalized regression analysis that can reduce the regression coefficient to 0 include a method using a regression equation (cost function) with an L1 norm regularization term, such as the Least Absolute Shrinkage and Selection Operator (Lasso), an elastic net, and SCAD (Smoothly Clipped Absolute Deviation). The results shown in FIG. 8, which will be described later, were obtained by using the Lasso.

[0071] When performing a regression analysis with penalties that can reduce the regression coefficient to 0, the corrosion-causing microorganism derivation unit 513 preferably uses the contents of all types of microbial groups derived based on the base sequence at the time of the preliminary test as explanatory variables (independent variables). However, this is not necessarily required. For example, if there is a microbial group that is known not to affect the corrosion of the metal material M, the corrosion-causing microorganism derivation unit 513 does not need to include the contents of that microbial group in the explanatory variables (independent variables).

[0072] Next, the corrosion-causing microorganism derivation unit 513 derives, as the corrosion-causing microorganism group, a microorganism group corresponding to a non-zero regression coefficient among the regression coefficients derived by the above-described penalized regression analysis that can reduce the regression coefficient to 0. The microorganism group corresponding to the regression coefficient is a microorganism group having a content represented by an explanatory variable (dependent variable) multiplied by the regression coefficient.

[0073] The results shown in Fig. 8, which will be described later, were obtained by using the regression coefficients derived as described above. The method for deriving the types of corrosion-causing microorganisms as an example of corrosion-causing microorganism attribute information as described above corresponds to the first selection step described in Japanese Patent No. 7299485.

[0074] As described in Japanese Patent No. 7299485, the corrosion-causing microorganism derivation unit 513 may further narrow down the types of microorganisms derived as described above.

[0075] For example, the corrosion-causing microorganism derivation unit 513 may create a set of B samples as the aforementioned samples. In this case, the corrosion-causing microorganism derivation unit 513 may derive regression coefficients using a penalized regression analysis that can reduce the regression coefficients to zero, and derive microbial groups corresponding to non-zero regression coefficients from the derived regression coefficients, for each of the set of B samples individually. The corrosion-causing microorganism derivation unit 513 may derive, for each type of microbial group, the occurrence frequency of the microbial group as U / B, where U is the number of times the microbial group is derived. In this case, the corrosion-causing microorganism derivation unit 513 may derive a microbial group whose occurrence frequency is equal to or greater than a predetermined value as the corrosion-causing microorganism group. The occurrence frequency corresponds to the reliability described in Japanese Patent No. 7299485.

[0076] Furthermore, for example, the corrosion-causing microorganism derivation unit 513 may perform a regression analysis using only the content of the microbial group, which is the explanatory variable (dependent variable) corresponding to a non-zero regression coefficient among the regression coefficients derived by a penalized regression analysis that can reduce the regression coefficient to 0, as the explanatory variable, and may perform a regression analysis using the natural potential of the metal material M as the objective variable, and derive the microbial group that shows either a positive correlation or a negative correlation as the corrosion-causing microorganism group.

[0077] Furthermore, for example, the corrosion-causing microorganism derivation unit 513 may perform a principal component regression analysis with a regularization term added, using only the content of the microorganism group, which is the explanatory variable (dependent variable) corresponding to a non-zero regression coefficient among the regression coefficients derived by a penalized regression analysis that can reduce the regression coefficient to 0, as the explanatory variable, and the spontaneous potential of the metal material M as the objective variable, and derive a microorganism group that shows at least either a positive correlation or a negative correlation as the corrosion-causing microorganism group.

[0078] Furthermore, for example, the corrosion-causing microorganism derivation unit 513 may perform a regression analysis using only the content of the microorganism group, which is the explanatory variable (dependent variable) corresponding to a non-zero regression coefficient among the regression coefficients derived by a penalized regression analysis that can reduce the regression coefficient to 0, as the explanatory variable, and the spontaneous potential of the metal material M as the objective variable, calculate the p-value and Akaike's information criterion, and derive the corrosion-causing microorganism group based on the results.

[0079] Furthermore, the corrosion-causing microorganism deriving unit 513 may derive the probability of each microorganism group affecting the corrosion of the metal material M (the probability that they may affect the corrosion of the metal material M) based on the results of narrowing down the microorganism groups as described above. For example, the corrosion-causing microorganism deriving unit 513 may determine that a microorganism group having an occurrence frequency equal to or greater than a predetermined value as a microorganism group having a relatively high probability of affecting the corrosion of the metal material M, and a microorganism group having an occurrence frequency less than the predetermined value as a microorganism group having a relatively low probability of affecting the corrosion of the metal material M. Furthermore, the corrosion-causing microorganism deriving unit 513 may determine that the higher the occurrence frequency of a microorganism group, the higher the probability that the microorganism group has an influence on the corrosion of the metal material M.

[0080] When the corrosion-causing microorganism attribute information derived by the corrosion-causing microorganism derivation unit 513 is stored in the storage unit 521 (i.e., when the processing of step S614 ends), the processing according to the flowchart shown in Fig. 6A ends. As described above, in this embodiment, a case where the third derivation means is realized by using the corrosion-causing microorganism derivation unit 513 is illustrated.

[0081] (Measurement / calculation results) FIG. 8 is a diagram showing an example of the measured value, total estimated value, and CV estimated value of the free-space potential of the metallic material M. The measured value is the measured value of the free-space potential of the metallic material M exemplified in FIG. 4 (note that the free-space potential values ​​used as a reference when making the free-space potential dimensionless are different between FIG. 4 and FIG. 8, so the free-space potential values ​​shown in FIG. 4 and FIG. 8 do not match). The total estimated value is the estimated value of the free-space potential of the metallic material M when a linear regression equation having regression coefficients derived by the corrosion-causing microorganism deriving unit 513 using all datasets is used as the estimation formula. The total estimated value is the value of the objective variable (free-space potential of the metallic material M) derived by substituting the explanatory variables (contents of corrosion-causing microorganisms) of all datasets into the linear regression equation. The CV estimated value is the estimated value of the free-space potential of the metallic material M derived by cross-validation. Specifically, the CV estimated value is the estimated value of the free-space potential of the metallic material M when a linear regression equation having regression coefficients derived by the corrosion-causing microorganism deriving unit 513 except for one dataset is used as the estimation formula. The CV estimated value is the value of the objective variable (the natural potential of the metal material M) derived by substituting the explanatory variable (the content of corrosion-causing microorganisms) of the excluded dataset into the linear regression equation, while changing the excluded dataset. Note that the total estimated value and the CV estimated value are arithmetic mean values.

[0082] The first dataset was created using data from days 3, 7, 14, 35, 60, and 64 after the start of the test. Figure 8(a) shows the results when the first dataset was used. The second dataset was created using data from days 3, 5, 7, 14, 35, 60, and 64 after the start of the test. Figure 8(b) shows the results when the second dataset was used. The difference between Figure 8(a) and Figure 8(b) is the presence or absence of data from day 5 after the start of the test.

[0083] Regardless of whether data was available from the fifth day after the test start date, the corrosion-causing microorganism derivation unit 513 derived the same two types of microbial groups (bacteria of the genus Solidesulfovibrio and bacteria of the genus Acidibacter) as types of microbial groups that may affect the corrosion of the metal material M (SUS304). The genus Solidesulfovibrio is a sulfate-reducing bacterium known to be involved in the corrosion of metal materials. On the other hand, no knowledge has been obtained as to whether bacteria of the genus Acidibacter are related to the corrosion of metal materials, but these results suggest that bacteria of the genus Acidibacter may have an effect on the corrosion of the metal material M (SUS304).

[0084] In addition, in Figure 8, r2 is the coefficient of determination R 2 8(a) and 8(b), it can be seen that by creating a data set without using the data on the fifth day, when the natural potential of metal material M is an outlier, it is possible to derive estimated values ​​(total estimated value and CV estimated value) that are closer to the measured value of metal material M, even if the number of time-series data of natural potential in the preliminary test and the number of time-series data of the content of microbial communities in the preliminary test are small (for example, the determination coefficient R obtained from the CV estimated value and the measured value shown in FIG. 8(a)). 2 = 0.985908, and the coefficient R obtained from the CV estimates and measurements shown in Figure 8(b) 2 =0.484744).

[0085] (Prediction acquisition unit 531, step S621) In step S621 of FIG. 6B, the prediction acquisition unit 531 acquires base sequence data at the time of prediction. In this embodiment, the time-series data of the base sequence at the time of prediction is time-series data of the base sequence of a microbial community contained in soil collected from the installation site or planned installation site of a metal material of the same type as the metal material M, as an example. The base sequence data at the time of prediction may be data at a single timing, or may be data at multiple timings (time-series data). The acquisition format of the base sequence data at the time of prediction may be the same as the acquisition format of the base sequence data at the time of pre-testing described above in the section (Pre-testing acquisition unit 511, step S611).

[0086] It should be noted that the processing according to the flowchart of Fig. 6A (storage of corrosion-causing microorganism attribute information by the storage unit 521) is completed before the processing of at least step S623 of the flowchart of Fig. 6B is started. For example, the processing according to the flowchart of Fig. 6B may be started after the processing according to the flowchart of Fig. 6A is completed.

[0087] 6B is executed when the above-described prediction is made (i.e., when corrosion prediction information regarding the possibility of corrosion of a metallic material is derived). Therefore, when executing the flowchart of FIG. 6B, it is preferable to use, as the sample S, a sample S having the same type of metallic material M as the metallic material from which the corrosion prediction information is to be derived.

[0088] When executing the flowchart of Fig. 6B, in step S611 of Fig. 6A, the pre-test acquisition unit 511 acquires, as time-series data of natural potential during a pre-test, time-series data of natural potential of the metal material M measured using the testing device 200 in which a sample S having the same type of metal material M as the metal material for which corrosion prediction information is to be derived is placed in a container 240. Also, in step S614 of Fig. 6A, the storage unit 521 stores, as corrosion-causing microorganism attribute information, attribute information of a microorganism group that may affect the corrosion of the same type of metal material M as the metal material for which corrosion prediction information is to be derived.

[0089] (Prediction time microorganism derivation unit 532, step S622) In step S622 of FIG. 6B, the microorganism derivation unit at prediction time 532 derives data on the content of the microorganism at prediction for each type of microorganism as an example of data of second microorganism community information based on the base sequence data of the microorganism at prediction. The base sequence data of the microorganism used by the microorganism at prediction time deriving unit 512 is data from the pre-test. In contrast, the base sequence data of the microorganism used by the microorganism at prediction time deriving unit 532 is data from the prediction. Furthermore, when the prediction acquisition unit 531 acquires data at a single timing as base sequence data at prediction, the microorganism at prediction time deriving unit 532 derives data at that timing for each type of microorganism as data on the content of the microorganism at prediction. Other processing by the microorganism at prediction time deriving unit 532 can be realized, for example, by the processing described above in the section (microorganism at prediction time deriving unit 512, step S612). Therefore, a detailed description of an example of the processing by the microorganism at prediction time deriving unit 532 will be omitted here. In addition, when the prediction acquisition unit 531 acquires time series data of the base sequence at the time of prediction, the prediction time microorganism derivation unit 532 may derive time series data of the content of the microbial group at the time of preliminary testing for each type of microbial group.

[0090] Like the first microbial community information, the second microbial community information may be information that can identify at least one of the quantity and quality of the microbial community. While the present embodiment illustrates a case in which the first microbial community information and the second microbial community information are the same type of information (specifically, content), they may also be different physical quantities. Furthermore, the prediction-time microbial derivation unit 532 does not necessarily need to derive the second microbial community information (content of the microbial community in this embodiment) as long as it has identified the microbial community contained in the soil at the installation site of the metal material. As described above, in this embodiment, a case where the fourth derivation means is realized by using the prediction time microorganism derivation unit 532 will be exemplified.

[0091] (Corrosion prediction information derivation unit 533, step S623) In step S623 of Fig. 6B, the corrosion prediction information derivation unit 533 derives corrosion prediction information based on the corrosion-causing microorganism attribute information stored in the storage unit 521. The corrosion-causing microorganism attribute information here is attribute information of a microorganism group that may affect the corrosion of the same type of metal material M as the metal material for which the corrosion prediction information is to be derived. In addition, the corrosion prediction information is information regarding the possibility of corrosion of the metal material at the installation location of the metal material.

[0092] For example, the corrosion prediction information derivation unit 533 may identify the types of microbial communities constituting the data on the content of microbial communities at the time of prediction derived for each type of microbial community by the prediction-time microbial derivation unit 532, and if the identified types of microbial communities include (at least one type of) a microbial community identified from the information stored in the storage unit 521, derive, as an example of corrosion prediction information, information indicating that the microbial communities contained in the soil at the installation site of the metal material M include a microbial community that may affect the corrosion of the metal material. In this way, the corrosion prediction information can indicate that the microbial communities contained in the soil at the installation site of the metal material include a microbial community that may affect the corrosion of the metal material.

[0093] Furthermore, the corrosion prediction information derivation unit 533 may include information (e.g., names) that can identify the types of microbial communities that may affect the corrosion of the metal material in the corrosion prediction information. In this way, the types of microbial communities that may affect the corrosion of the metal material can be specifically indicated in the corrosion prediction information.

[0094] Furthermore, in this embodiment, an example is given in which the corrosion prediction information deriving unit 533 further uses the content of the microbial community at the time of prediction (second microbial community information) derived for each type of microbial community by the prediction-time microbial community deriving unit 532 when deriving the corrosion prediction information. In this case, if a microbial community identified from the corrosion-causing microbial community attribute information stored in the storage unit 521 is included among the types of microbial communities identified from the content of the microbial community at the time of prediction for each type of microbial community, the corrosion prediction information deriving unit 533 may derive, as an example of corrosion prediction information, information indicating the degree of corrosion possibility of the metallic material at the installation location or planned installation location, or the presence or absence of signs of corrosion of the metallic material at the installation location, according to the content of the microbial community (second microbial community information). In this way, the degree of corrosion possibility of the metallic material and the presence or absence of signs of corrosion of the metallic material can be indicated by the corrosion prediction information.

[0095] For example, when a type of microbial group identified from the content of the microbial group at the time of prediction for each type of microbial group includes a microbial group identified from the information stored in the memory unit 521, the corrosion prediction information derivation unit 533 may determine whether the content of the microbial group that may affect the corrosion of the metal material exceeds a predetermined value.

[0096] In this case, the corrosion prediction information derivation unit 533 may derive, as an example of corrosion prediction information, information indicating a relatively high possibility of corrosion of the metal material when the content of microbial communities that can affect the corrosion of the metal material exceeds a predetermined value. Otherwise, the corrosion prediction information derivation unit 533 may derive, as an example of corrosion prediction information, information indicating a relatively low possibility of corrosion of the metal material. In this way, for example, if there is a location where an object (e.g., a metal pipe) containing the metal material for which corrosion prediction information is to be derived is planned to be installed and the content of microbial communities that can affect the corrosion of the metal material exceeds a predetermined value, the corrosion prediction information can indicate that there is a relatively high possibility of corrosion of the metal material if the metal material is installed in that location. Furthermore, for example, if an object (e.g., a metal pipe) containing the metal material for which corrosion prediction information is to be derived is installed and the content of microbial communities that can affect the corrosion of the metal material exceeds a predetermined value, the corrosion prediction information can indicate that there are signs of corrosion of the metal material. Note that the content of microbial communities that can affect the corrosion of the metal material may be the content of each type of microbial community or the total content of all types of microbial communities that can affect the corrosion of the metal material.

[0097] Furthermore, for example, when the prediction time microorganism derivation unit 532 derives time series data of the content of the microbial group at the time of prediction, and the type of microbial group identified from the content of the microbial group at the time of prediction for each type of microbial group includes a microbial group identified from the information stored in the memory unit 521, the corrosion prediction information derivation unit 533 may determine whether the absolute value of the change in the content of the microbial group per unit time exceeds a predetermined value. In this case, for example, when the absolute value of the change per unit time in the content of the microbial community (a microbial community that can affect the corrosion of the metal material) exceeds a predetermined value, the corrosion prediction information derivation unit 533 may derive, as an example of corrosion prediction information, information indicating a sudden increase in the possibility of corrosion of the metal material. Note that in this case as well, the content of the microbial community (a microbial community that can affect the corrosion of the metal material) may be the content of each type of microbial community, or may be the total content of all types of microbial communities that can affect the corrosion of the metal material.

[0098] Furthermore, for example, the corrosion prediction information derivation unit 533 may derive corrosion prediction information based on both the change per unit time in the content of the microbial group (a microbial group that may affect the corrosion of the metal material) and the content of the microbial group. In this case, for example, when the absolute value of the change per unit time in the content of the microbial community (a microbial community that can affect the corrosion of the metal material) exceeds a predetermined value and the content of the microbial community also exceeds a predetermined value, the corrosion prediction information derivation unit 533 may derive, as corrosion prediction information, information indicating that there are signs of corrosion of the metal material at the installation location. Note that, in this case as well, the content of the microbial community may be the content of each type of microbial community, or may be the total content of all types of microbial communities that can affect the corrosion of the metal material.

[0099] In addition, the corrosion prediction information derivation unit 533 may derive, as an example of corrosion prediction information, information that indicates in stages the degree of possibility of corrosion of the metal material or the degree of signs of corrosion of the metal material by using multiple predetermined values ​​as predetermined values ​​to compare with the content (size) of the aforementioned microbial group or the unit time of the content of the microbial group. In this embodiment, the corrosion prediction information derivation unit 533 is used to realize an example of a fifth derivation means.

[0100] (Output unit 541, step S624) 6B, the output unit 541 outputs the corrosion prediction information derived by the corrosion prediction information derivation unit 533. The output unit 541 may perform at least one of, for example, displaying the corrosion prediction information on a computer display, transmitting the corrosion prediction information to an external device, and storing the corrosion prediction information in a storage medium. When the corrosion prediction information is output by the output unit 541 (that is, when the processing of step S624 ends), the processing according to the flowchart shown in FIG. 6B ends.

[0101] (summary) As described above, in this embodiment, the corrosion factor detection system performs measurements on a sample S including a test liquid L containing a microorganism population and a metal material M in contact with the test liquid L, thereby deriving a corrosion process-reflecting physical quantity, which is a physical quantity whose value changes depending on the corrosion process of the metal material M. The corrosion factor detection system also derives first microbial community information, which is information that can identify at least one of the quantity and quality of the microorganism population contained in the test liquid L. The corrosion factor detection system then derives corrosion-causing microorganism attribute information based on the time-series data of the corrosion process-reflecting physical quantity and the time-series data of the first microbial community information. Therefore, it is possible to identify the attributes of the microorganism population that affect the corrosion of the metal material.

[0102] Furthermore, in this embodiment, the corrosion factor detection system derives the physical quantities reflecting the corrosion process of the metal material M in contact with the test solution L containing a microbial population at the installation location or planned installation location of a metal material of the same type as the metal material M. Therefore, it is possible to identify the attributes of the microbial population present at the installation location or planned installation location of the metal material as a microbial population that may affect the corrosion of the metal material.

[0103] Furthermore, in this embodiment, the corrosion factor detection system derives attribute information of a microbial community at an installation location or planned installation location of a metallic material of the same type as the metallic material M. Then, the corrosion factor detection system derives corrosion prediction information regarding the possibility of corrosion of the metallic material at the installation location or planned installation location based on the attribute information of the microbial community and the corrosion-causing microorganism attribute information described above. Therefore, objective information can be obtained as information regarding the possibility of corrosion of the metallic material at the installation location or planned installation location of the metallic material.

[0104] Furthermore, in this embodiment, the corrosion factor detection system derives information (second microbial community information) capable of identifying at least one of the quantity and quality of microbial communities collected from the installation location or planned installation location of a metallic material of the same type as metallic material M. The corrosion factor detection system then derives corrosion prediction information based on the information (second microbial community information) capable of identifying at least one of the quantity and quality of the microbial communities and the attribute information of the microbial communities that may affect the corrosion of the metallic material. Therefore, more detailed information (e.g., not only whether corrosion is possible but also the degree of corrosion possibility and whether signs of corrosion are present) can be more easily derived as information (corrosion prediction information) regarding the possibility of corrosion of the metallic material of the same type as metallic material M at the installation location or planned installation location. In this case, if the corrosion factor detection system derives each piece of information (physical quantity) capable of identifying the quantity of microbial communities collected at multiple times from the installation location or planned installation location of a metallic material of the same type as metallic material M, it can derive the time-dependent changes in the quantity and quality of microbial communities that may affect the corrosion of the metallic material at the installation location or planned installation location. This makes it possible to obtain more diverse information, for example, about the microorganisms that may affect the corrosion of the metal material.

[0105] Furthermore, in this embodiment, the corrosion factor detection system uses the natural potential of the metallic material M as the physical quantity reflecting the corrosion process. Therefore, it is possible to use a physical quantity that indicates the state of the metallic material M in the corrosion process from the viewpoint of the quality of the metallic material M. Furthermore, it is possible to measure the physical quantity reflecting the corrosion process relatively easily.

[0106] Furthermore, in this embodiment, the corrosion factor detection system uses a machine learning model that has the corrosion process-reflecting physical quantity as a response variable and the first microbial community information as an explanatory variable to select the first microbial community information corresponding to the corrosion process-reflecting physical quantity, and derives a microbial community that may affect the corrosion of the metal material M based on the selection result. Therefore, the microbial community that may affect the corrosion of the metal material M can be derived without using a rule-based method or a physical model (e.g., a differential equation that describes a physical phenomenon). Therefore, the microbial community that may affect the corrosion of the metal material M can be derived without manually setting rules or describing the corrosion process of the metal material M with a mathematical formula.

[0107] In this embodiment, the corrosion factor detection system uses a regression equation as the machine learning model. The corrosion factor detection system derives regression coefficients of the regression equation based on time-series data of the corrosion process-reflecting physical quantity and time-series data of the first microbial community information, selects the first microbial community information corresponding to the corrosion process-reflecting physical quantity based on the derived results, and derives a microbial community that may affect the corrosion of the metal material M based on the selected results. Therefore, for example, the basis for deriving a microbial community that may affect the corrosion of the metal material M can be understood based on the regression coefficients of the regression equation. Furthermore, by deriving the regression coefficients of the regression equation by performing a penalized regression analysis that can reduce the regression coefficients to 0, for example, the microbial community that may affect the corrosion of the metal material M can be derived with higher accuracy.

[0108] Furthermore, in this embodiment, the corrosion factor detection system creates a sample by resampling using the time series data of the corrosion process-reflecting physical quantity and the time series data of the first microbial community information, and uses the created sample to derive a microbial community that may affect the corrosion of the metal material M. Therefore, even if there is a small amount of time series data of the corrosion process-reflecting physical quantity and the first microbial community information, it is possible to suppress a decrease in the accuracy of deriving a microbial community that may affect the corrosion of the metal material M.

[0109] Furthermore, in this embodiment, the corrosion factor detection system derives, as the first microbial community information, information that can identify the amount of microbial community for each type of microbial community, and derives, as the corrosion-causing microbial community attribute information, information that can identify the type of microbial community that can affect the corrosion of the metal material. Therefore, it is possible to obtain information about what type of microbial community can affect the corrosion of the metal material.

[0110] [Second embodiment] Next, a second embodiment will be described. In the first embodiment, the first microbial community information is information that can identify the amount of microbial community for each type of microbial community, and the attribute information of microbial community that can affect the corrosion of metal materials is information that can identify the type of microbial community that can affect the corrosion of metal materials. In contrast, the present embodiment exemplifies a case where the first microbial community information is information that can identify the function of the microbial community, and the attribute information of microbial community that can affect the corrosion of metal materials is information that can identify the function of the microbial community that can affect the corrosion of metal materials. Furthermore, the present embodiment exemplifies a case where the second microbial community information, like the first microbial community information, is information that can identify the function of the microbial community. Thus, the present embodiment differs from the first embodiment mainly in the configuration and processing due to differences in the first microbial community information, the second microbial community information, and the attribute information of microbial community that can affect the corrosion of metal materials. Therefore, in the description of this embodiment, the same parts as those in the first embodiment are designated by the same reference numerals as those in FIGS. 1 to 8, and detailed description thereof will be omitted.

[0111] The function of a microbial community can be identified, for example, by the function of the genes possessed by the microbial community. Therefore, this embodiment illustrates a case where the first microbial community information and the second microbial community information are information capable of identifying the amount of microbial communities that have genes with the same function. The amount of microbial communities that have genes with the same function may be an absolute amount (the amount itself) or a relative amount (relative ratio). Note that the amount of microbial communities that have genes with the same function is determined for each type of gene function.

[0112] Furthermore, in this embodiment, an example is given in which the attribute information of a microbial community that may affect the corrosion of a metal material is information that can identify the functions of genes contained in the microbial community that may affect the corrosion of the metal material. As described in the first embodiment, cases in which a microbial community may affect the corrosion of a metal material may include cases in which a microbial community clearly affects the corrosion of the metal material M, as well as cases in which a microbial community may potentially affect the corrosion of the metal material M.

[0113] Furthermore, this embodiment illustrates a case where the functions of genes possessed by a microbial community are derived for each type of gene function based on the base sequence (e.g., ribosomal RNA gene sequence) of the microbial community obtained by analysis using the sequencer 120. As described in the first embodiment, during a preliminary test, for example, the base sequence of a microbial community contained in a test solution L is derived by the sequencer 120. During prediction, the sequencer 120 derives the base sequence of a microbial community contained in the soil at the installation site and the planned installation site of a metal material of the same type as the metal material M used during the preliminary test.

[0114] The method for deriving the functions of genes possessed by microbial communities contained in test solution L or soil based on the base sequences of the microbial communities contained in test solution L or soil can be realized using known techniques and is not limited thereto. However, in this embodiment, the type of microbial community and the amount of each microbial community are derived based on the base sequences of the microbial community contained in test solution L or soil, and the type of gene function possessed by the microbial community and the amount of each gene obtained from the amount of microbial communities having genes with the same function are derived based on the derived nucleotide sequences. As an example of such a method, a method may be used in which information obtained using software (PICRUSt2) described in Non-Patent Document 1 is mapped to a database (KO (KEGG ORTHOLOGY) Database) described in Non-Patent Document 2. When using PICRUSt2, the amount of microbial communities having genes with the same function is obtained as abundance. Furthermore, in PICRUSt2, multiple types of genes classified into the same KO (KEGG ORTHOLOGY) are considered to have the same function. In the following description, a group of genes having the same function will be referred to as a group of genes with the same function as necessary.

[0115] The corrosion factor detection system and corrosion factor detection method of the present embodiment differ from the corrosion factor detection system and corrosion factor detection method of the first embodiment in some respects, including the preliminary test microorganism derivation unit 512 and step S612, the corrosion-causing microorganism derivation unit 513, the memory unit 521, and steps S613 and S614, the prediction acquisition unit 531 and step S621, and the prediction-time microorganism derivation unit 532 and step S622. Therefore, the following describes the preliminary test microorganism derivation unit 512 and step S612, the corrosion-causing microorganism derivation unit 513, the memory unit 521, and steps S613 and S614, the prediction acquisition unit 531 and step S621, and the prediction-time microorganism derivation unit 532 and step S622, focusing on the differences from the first embodiment.

[0116] In this embodiment, in FIG. 5, the "content of microbial community (first microbial community information)" shown next to the arrow pointing from the pre-test microorganism deriving unit 512 to the corrosion-causing microorganism deriving unit 513 is replaced with the "amount of the same functional gene group (first microbial community information)." Furthermore, the "content of microbial community (second microbial community information)" shown next to the arrow pointing from the prediction microorganism deriving unit 532 to the corrosion prediction information deriving unit 533 is replaced with the "amount of the same functional gene group (second microbial community information)." Furthermore, the "content of microbial community (first microbial community information)" in the block indicated by S612 in FIG. 6A is replaced with the "amount of the same functional gene group (first microbial community information)." Furthermore, the "content of microbial community (second microbial community information)" in the block indicated by S622 in FIG. 6A is replaced with the "amount of the same functional gene group (second microbial community information)."

[0117] (Pre-test microorganism extraction unit 512, step S612) In step S612 of Figure 6A, first, the pre-test microorganism derivation unit 512 derives time series data of the content of the microorganism group at the time of the pre-test for each type of microorganism group based on time series data of the base sequence (ribosomal RNA gene sequence) of the microorganism group at the time of the pre-test, as described in the first embodiment.

[0118] Based on the time series data of the content of the microbial group at the time of the pre-test derived in this manner, the pre-test microorganism derivation unit 512 derives the data at the time of the pre-test as time series data of the amount of the same functional gene group for each type of gene function.

[0119] In this embodiment, an example is given in which the pre-test microorganism derivation unit 512 analyzes time series data of ribosomal RNA gene sequences during the pre-test using PICRUSt2, thereby deriving time series data of the amount (abundance) of genes with the same function during the pre-test for each type of gene function.

[0120] As described above, this embodiment illustrates a case in which the second derivation means is realized by using the preliminary test microorganism derivation unit 512. Furthermore, this embodiment makes it possible to identify the types of genes possessed by a microorganism group and the function of each type of gene from the amount of the same functional gene group. In this way, this embodiment illustrates a case in which the first microorganism group information is information that can identify the quality of a microorganism group.

[0121] (Corrosion-causing microorganism deriving unit 513, memory unit 521, steps S613 and S614) The processing in steps S613 and S614 of Figure 6A in this embodiment is realized by replacing the time series data of the content of the microbial group during the preliminary test described in the first embodiment with time series data of the amount of the same functional gene group during the preliminary test in the description of the processing in steps S613 and S614 in the first embodiment.

[0122] That is, the corrosion-causing microorganism derivation unit 513 creates a dataset including data obtained at the same time, for example, data on the natural potential during the preliminary test and data on the amount of the same functional gene group during the preliminary test, and by using the created dataset, derives the corrosion-causing microorganism group by the method described in Patent Publication No. 7299485.

[0123] In the first embodiment, the corrosion-causing microorganism group includes one type of microorganism group, and the type of the microorganism group is corrosion-causing microorganism attribute information (information that can identify the type of microorganism group that can affect the corrosion of metal materials). In contrast, in the present embodiment, the corrosion-causing microorganism group can include multiple types of microorganism groups that have the same functions. Furthermore, in the present embodiment, the type of function identified from the corrosion-causing microorganism group (the type of the same function) is corrosion-causing microorganism attribute information. When there are multiple types of corrosion-causing microorganism group, multiple types of gene functions are derived as an example of corrosion-causing microorganism attribute information.

[0124] Then, in step S614, the storage unit 521 stores the corrosion-causing microorganism attribute information derived by the corrosion-causing microorganism derivation unit 513. When the corrosion-causing microorganism attribute information derived by the corrosion-causing microorganism derivation unit 513 is stored in the storage unit 521 (i.e., when the processing of step S614 ends), the processing according to the flowchart shown in Fig. 6A ends. As described above, in this embodiment, a case where the third derivation means is realized by using the corrosion-causing microorganism derivation unit 513 is illustrated.

[0125] (Measurement / calculation results) The ribosomal RNA gene sequence data of the microbial community contained in test solution L, obtained by the sequencer 120 when creating the data set shown in (Measurement and calculation results) of the first embodiment, was collected on days 0, 3, 5, 6, 14, 35, 60, and 64 from the start of the test. The ribosomal RNA gene sequence data of the microbial community contained in test solution L was analyzed using PICRUSt2 (and the KO (KEGG ORTHOLOGY) Database) to estimate the functions of the genes possessed by the microbial community. As a result, the existence of 9,374 types of gene functions was discovered.

[0126] Data sets were created using the same data as those shown in (Calculation Results) in the first embodiment. Here, a third data set was created using data from the 5th, 7th, 14th, 35th, 60th, and 64th days after the start of the test. A fourth data set was created using data from the 0th, 7th, 14th, 35th, 60th, and 64th days after the start of the test.

[0127] As a result of narrowing down the gene functions as described in this embodiment (deriving a group of genes with the same function that constitutes the corrosion-causing microorganism group), in the third dataset, of the 9,374 gene functions mentioned above, K00799, K02529, K02014, and K14266 were estimated to be the functions of genes whose frequency changes with time in the natural potential, and in the fourth dataset, K02014, K00799, K07165, and K02529 were estimated to be the functions of genes whose frequency changes with time in the natural potential. Note that K00799 and the like are KO (KEGG ORTHOLOGY) identification information (ID).

[0128] FIG. 9 shows an example of the measured value, total estimated value, and CV estimated value of the half-cell potential of metal material M. The measured value is the measured value of the half-cell potential of metal material M exemplified in FIG. 4. The total estimated value and CV estimated value are derived in the same manner as the total estimated value and CV estimated value shown in FIG. 8. However, as described above, in this embodiment, the corrosion-causing microorganism derivation unit 513 uses a linear regression equation in which the amount of the same functional gene group is used as the explanatory variable, rather than the content of the microbial group. FIG. 9(a) shows the results when the third dataset is used, and FIG. 9(b) shows the results when the fourth dataset is used. The difference between FIG. 9(a) and FIG. 9(b) is that FIG. 9(a) uses data from day 0 after the start of the test, while FIG. 9(b) uses data from day 7 after the start of the test.

[0129] Also, in Figure 9, as in Figure 8, r2 is the coefficient of determination R 2 As shown in Figures 9(a) and 9(b), even if the number of time-series data of the spontaneous potential during the preliminary test and the number of time-series data of the content of the microbial community during the preliminary test are small, it is possible to derive estimated values ​​(total estimated value and CV estimated value) that are closer to the measured value of the metal material M (for example, the determination coefficient R obtained from the CV estimated value and the measured value shown in Figure 9(a)). 2 = 0.958326, and the coefficient R obtained from the CV estimates and measurements shown in Figure 9(b) 2 =0.922608).

[0130] Table 1 shows the entry (#KO_ID) of the gene function (KO (KEGG ORTHOLOGY)) narrowed down as described above, the symbol and name of the gene identified by the ID (see "Corresponding gene and enzyme name"), and a summary of the function of the gene (see "Remarks" column). In Table 1, the symbols and names are shown to the left of the ";" in the "Corresponding gene and enzyme name" column. The entries, symbols, and names can be searched for in Non-Patent Document 2.

[0131] [Table 1]

[0132] In Table 1, K02014 is a function related to iron complexes (siderophores), and K07165 is a function related to iron transport. These functions are thought to be related to the corrosion of metallic materials. The exact relationship between the other functions and the corrosion of metallic materials is unclear, but they may be functions related to unknown corrosion mechanisms.

[0133] (Prediction time microorganism derivation unit 532, step S622) In the process of step S612 in FIG. 6A, data from the preliminary test is used, and time-series data of the content of the microbial community is derived for each type of microbial community. On the other hand, in the process of step S622 in FIG. 6B, data from the time of prediction is used. Also, in the process of step S622 in FIG. 6B, the data of the content of the microbial community may be data at a single timing, or data at multiple timings (time-series data). As described in the first embodiment, if the prediction acquisition unit 531 acquires data at a single timing as base sequence data at the time of prediction, the data of the content of the microbial community may be data at a single timing. Also, if the prediction acquisition unit 531 acquires data at multiple timings as base sequence data at the time of prediction, the data of the content of the microbial community may be time-series data. In these respects, the process of step S622 in FIG. 6B differs from step S612 in FIG. 6A.

[0134] As described in the first embodiment, the microorganism derivation unit 532 at prediction time derives the content of the microorganism at prediction time for each type of microorganism based on the data of the base sequences (ribosomal RNA gene sequences) of the microorganism at prediction time. Then, the microorganism derivation unit 512 at pre-test time derives data on the amount of the homofunctional gene group at prediction time for each type of gene function based on the content of the microorganism at prediction time. This embodiment illustrates a case where the fourth derivation means is realized by using the prediction-time microorganism derivation unit 532. This embodiment also illustrates a case where the amount of the same functional gene group at the time of prediction is the second microbial community information.

[0135] (Corrosion prediction information derivation unit 533, step S623) In step S623 of FIG. 6B, the corrosion prediction information deriving unit 533 derives corrosion prediction information based on the corrosion-causing microorganism attribute information stored in the storage unit 521.

[0136] In this embodiment, for example, if the functions identified from the group of identically functional genes at the time of prediction (groups of multiple types of microorganisms having the same function) derived by the microorganism derivation unit 532 at the time of prediction include (at least one type of) the function of a gene identified from the information stored in the memory unit 521, the corrosion prediction information derivation unit 533 may determine that the group of identically functional genes is a group of identically functional genes that can affect the corrosion of the metal material, and may derive information indicating that the functions possessed by the group of microorganisms contained in the soil at the installation site of the metal material M include a function that can affect the corrosion of the metal material, as an example of corrosion prediction information.

[0137] Furthermore, the corrosion prediction information derivation unit 533 may include information (for example, a name) that can identify the type of gene that can affect the corrosion of the metal material in the corrosion prediction information.

[0138] Furthermore, this embodiment illustrates a case where, when deriving corrosion prediction information, the corrosion prediction information derivation unit 533 further uses the amount of the same functional gene group at the time of prediction (second microbial community information) derived by the microorganism derivation unit 532 at the time of prediction. In this case, if the functions of genes identified from the same functional gene group at the time of prediction include a function identified from the corrosion-causing microorganism attribute information stored in the storage unit 521, the corrosion prediction information derivation unit 533 may derive, as an example of corrosion prediction information, information indicating the degree of possibility of corrosion of the metallic material at the installation location or planned installation location, or the presence or absence of signs of corrosion of the metallic material at the installation location of the metallic material, depending on the amount of the same functional gene group.

[0139] For example, if the functions of genes identified from the group of genes with the same function at the time of prediction include a function identified from the information stored in the memory unit 521, the corrosion prediction information derivation unit 533 may determine whether the amount of the group of genes with the same function exceeds a predetermined value.

[0140] In this case, the corrosion prediction information derivation unit 533 may derive, as an example of corrosion prediction information, information indicating a relatively high possibility that the metal material will corrode when the amount of the same-function gene group that can affect the corrosion of the metal material exceeds a predetermined value, and may derive, as an example of corrosion prediction information, information indicating a relatively low possibility that the metal material will corrode when this amount does not exceed a predetermined value. Note that the amount of the same-function gene group that can affect the corrosion of the metal material may be the amount for each same-function gene group, or may be the total amount of all same-function gene groups that can affect the corrosion of the metal material.

[0141] Furthermore, for example, when the prediction time microorganism derivation unit 532 derives time series data on the amount of the same functional gene group at the time of prediction, and the functions of the genes identified from the same functional gene group at the time of prediction include a function identified from the information stored in the memory unit 521, the corrosion prediction information derivation unit 533 may determine whether the absolute value of the change in the amount of the same functional gene group per unit time exceeds a predetermined value. In this case, for example, when the absolute value of the change per unit time of the amount of the same-function gene group at the time of prediction exceeds a predetermined value, the corrosion prediction information derivation unit 533 may derive, as an example of corrosion prediction information, information indicating a sudden increase in the possibility of corrosion of the metal material. Note that in this case as well, the amount of the same-function gene group at the time of prediction (the same-function gene group that can affect the corrosion of the metal material) may be the amount for each same-function gene group, or may be the total value of the amounts of all the same-function gene groups that can affect the corrosion of the metal material.

[0142] Furthermore, for example, the corrosion prediction information derivation unit 533 may derive corrosion prediction information based on both the change per unit time in the amount of the same functional gene group (the same functional gene group that can affect the corrosion of the metal material) and the amount of the same functional gene group. In this case, for example, when the absolute value of the change per unit time in the amount of the same functional gene group (the same functional gene group that can affect the corrosion of the metal material) exceeds a predetermined value and the amount of the same functional gene group also exceeds a predetermined value, the corrosion prediction information derivation unit 533 may derive, as corrosion prediction information, information indicating that there are signs of corrosion of the metal material at the installation location. Note that in this case as well, the amount of the same functional gene group (may be the amount for each same functional gene group, or may be the total amount of all same functional gene groups that can affect the corrosion of the metal material.

[0143] In addition, the corrosion prediction information derivation unit 533 may use multiple predetermined values ​​as predetermined values ​​to compare with the amount (size) of the aforementioned homofunctional microorganism group or the amount of the homofunctional gene group per unit time, to derive information that indicates in stages the degree of possibility of corrosion of the metal material or the degree of signs of corrosion of the metal material, as an example of corrosion prediction information. In this embodiment, the corrosion prediction information derivation unit 533 is used to realize an example of a fifth derivation means.

[0144] (summary) As described above, in this embodiment, the corrosion factor detection system derives, as the first microbial community information, information that can identify the function of a microbial community, and derives, as the corrosion-causing microbial attribute information, information that can identify the function of a microbial community that can affect the corrosion of a metal material. Therefore, information can be obtained about what function of a microbial community can affect the corrosion of a metal material. It should be noted that the present embodiment can also employ the various modifications described in the first embodiment.

[0145] [Other embodiments] The above-described embodiments of the present disclosure can be realized by a computer executing a program. A computer-readable recording medium on which the program is recorded and a computer program product such as the program can also be applied as embodiments of the present disclosure. Examples of recording media that can be used include flexible disks, hard disks, optical disks, magneto-optical disks, CD-ROMs, magnetic tapes, non-volatile memory cards, and ROMs. The embodiments of the present disclosure may also be realized by dedicated hardware such as an ASIC (Application Specific Integrated Circuit). Furthermore, the above-described embodiments of the present disclosure are merely examples of specific embodiments for carrying out the present disclosure, and the technical scope of the present disclosure should not be interpreted as being limited by these. In other words, the present disclosure can be embodied in various forms without departing from its technical concept or main features.

[0146] The disclosure of the above embodiment can be implemented as follows, for example. [Disclosure 1] a first derivation means for deriving a corrosion process reflection physical quantity, which is a physical quantity whose value changes according to the corrosion process of the metal material, by measuring a sample including a medium containing a microorganism group and a metal material in contact with the medium; a second derivation means for deriving first microbial community information, which is information that can identify at least one of the quantity and quality of the microbial community contained in the medium; a third derivation means for deriving attribute information of a microbial community that may affect the corrosion of the metal material based on the time series data of the corrosion process-reflecting physical quantity derived by the first derivation means and the time series data of the first microbial community information derived by the second derivation means; A corrosion factor detection system comprising: [Disclosure 2] The corrosion factor detection system described in Disclosure 1, wherein the medium includes a microbial population at a location where a metal material of the same type as the metal material is installed, or a microbial population at a planned location where a metal material of the same type as the metal material is installed. [Disclosure 3] A fourth derivation means for deriving attribute information of a microorganism group present at an installation location of a metal material of the same type as the metal material, or attribute information of a microorganism group present at a planned installation location of a metal material of the same type as the metal material; The corrosion factor detection system according to Disclosure 1 or 2, further comprising a fifth derivation means for deriving corrosion prediction information regarding the possibility of corrosion of the metal material at the installation location or the planned installation location, based on attribute information of a microbial group that may affect the corrosion of the metal material, derived by the third derivation means, and attribute information of the microbial group, derived by the fourth derivation means. [Disclosure 4] the fourth derivation means further derives second microbial community information, which is information representing at least one of the quantity and quality of the microbial community collected from the installation site or the planned installation site; The corrosion factor detection system described in Disclosure 3, wherein the fifth derivation means derives the corrosion prediction information based on attribute information of a microbial group that may affect the corrosion of the metal material derived by the third derivation means and the second microbial group information of the microbial group derived by the fourth derivation means. [Disclosure 5] The corrosion factor detection system described in Disclosure 4, wherein the fourth derivation means derives as the second microbial group information each piece of information capable of identifying the type or quality of the microbial group collected at multiple times from the installation location or the planned installation location. [Disclosure 6] 6. The corrosion factor detection system according to any one of claims 1 to 5, wherein the physical quantity reflecting the corrosion process includes a natural potential. [Disclosure 7] The corrosion factor detection system according to any one of Disclosures 1 to 6, wherein the third derivation means selects the first microbial community information corresponding to the corrosion process-reflecting physical quantity based on the time series data of the corrosion process-reflecting physical quantity and the time series data of the first microbial community information using a machine learning model having the corrosion process-reflecting physical quantity as a target variable and the first microbial community information as an explanatory variable, and derives attribute information of a microbial community that may affect the corrosion of the metal material based on the selection result. [Disclosure 8] the machine learning model includes a regression equation, The corrosion factor detection system described in Disclosure 7, wherein the third derivation means derives a regression coefficient of the regression equation based on time series data of the corrosion process-reflecting physical quantity and time series data of the first microbial community information, selects the first microbial community information corresponding to the corrosion process-reflecting physical quantity based on the derived result, and derives attribute information of a microbial community that may affect the corrosion of the metal material based on the selected result. [Disclosure 9] The corrosion factor detection system described in Disclosure 8, wherein the third derivation means derives the regression coefficient of the regression equation by performing a penalized regression analysis that can reduce the regression coefficient to 0 based on the time series data of the physical quantity reflecting the corrosion process and the time series data of the first microbial community information. [Disclosure 10] The corrosion factor detection system according to any one of Disclosures 7 to 9, wherein the third derivation means creates a sample by resampling using the time series data of the physical quantity reflecting the corrosion process and the time series data of the first microbial community information, and derives attribute information of the microbial community that may affect the corrosion of the metal material using the created sample. [Disclosure 11] the first microbial community information is information that can identify the amount of the microbial community for each type of the microbial community, A corrosion factor detection system described in any one of Disclosures 1 to 10, wherein the attribute information of the microbial group that may affect the corrosion of the metal material is information that can identify the type of microbial group that may affect the corrosion of the metal material. [Disclosure 12] the first microbial community information is information that can identify the function of the microbial community, A corrosion factor detection system described in any one of Disclosures 1 to 10, wherein the attribute information of the microbial group that may affect the corrosion of the metal material is information that can identify a function among the functions of the microbial group that may affect the corrosion of the metal material. [Disclosure 13] 13. The corrosion factor detection system according to any one of Disclosures 1 to 12, wherein the microbial community includes a microbial species derived from a ribosomal RNA gene. [Disclosure 14] a first derivation step of deriving a corrosion process-reflecting physical quantity, which is a physical quantity whose value changes depending on the corrosion process of the metal material, by measuring a sample including a medium containing a microorganism group and a metal material in contact with the medium; a second derivation step of deriving first microbial community information, which is information that can identify at least one of the quantity and quality of the microbial community contained in the medium; a third derivation step of deriving attribute information of a microbial community that may affect the corrosion of the metal material based on the time series data of the corrosion process-reflecting physical quantity derived in the first derivation step and the time series data of the first microbial community information derived in the second derivation step; A corrosion factor detection method comprising: [Disclosure 15] 14. A program for causing a computer to function as each means of the corrosion factor detection system according to any one of Disclosures 1 to 13. [Explanation of symbols]

[0147] 110 DNA extraction device 120 Sequencer 130 Potential measuring device 140 Information processing equipment 200 Test Equipment 210 lead wire 220 tube 230 Reference electrode 240 Container 250 stopper 260 lead wire 511 Pre-examination acquisition section 512 Microorganism extraction section during pre-test 513 Corrosion factor microorganism extraction section 521 Storage section 531 Prediction acquisition unit 532 Microbial extraction part at the time of prediction 533 Corrosion prediction information derivation part 541 Output section L test solution M Metal material S sample

Claims

1. a first derivation means for deriving a corrosion process-reflecting physical quantity, which is a physical quantity whose value changes depending on the corrosion process of the metal material, by measuring a sample including a medium containing a microorganism group and a metal material in contact with the medium; a second derivation means for deriving first microbial community information, which is information that can identify at least one of the quantity and quality of the microbial community contained in the medium; a third derivation means for deriving attribute information of a microbial community that may affect the corrosion of the metal material based on the time series data of the corrosion process-reflecting physical quantity derived by the first derivation means and the time series data of the first microbial community information derived by the second derivation means; A corrosion factor detection system comprising:

2. The corrosion factor detection system according to claim 1 , wherein the medium includes a microbial population at a location where a metal material of the same type as the metal material is installed, or a microbial population at a planned location where a metal material of the same type as the metal material is to be installed.

3. a fourth derivation means for deriving attribute information of a microorganism group present at an installation location of a metal material of the same type as the metal material, or attribute information of a microorganism group present at a planned installation location of a metal material of the same type as the metal material; 3. The corrosion factor detection system according to claim 1, further comprising a fifth derivation means for deriving corrosion prediction information regarding the possibility of corrosion of the metal material at the installation location or the planned installation location based on attribute information of a microbial group that may affect the corrosion of the metal material derived by the third derivation means and attribute information of the microbial group derived by the fourth derivation means.

4. the fourth derivation means further derives second microbial community information, which is information representing at least one of the quantity and quality of the microbial community collected from the installation site or the planned installation site; The corrosion factor detection system of claim 3, wherein the fifth derivation means derives the corrosion prediction information based on attribute information of a microbial group that may affect the corrosion of the metal material derived by the third derivation means and the second microbial group information of the microbial group derived by the fourth derivation means.

5. The corrosion factor detection system of claim 4, wherein the fourth derivation means derives as the second microbial group information each piece of information capable of identifying the quality of the microbial group collected at multiple times from the installation location or the planned installation location.

6. The corrosion factor detection system according to claim 1 , wherein the physical quantity reflecting the corrosion process includes a natural potential.

7. The corrosion factor detection system according to claim 1 or 2, wherein the third derivation means selects the first microbial community information corresponding to the corrosion process-reflecting physical quantity based on the time series data of the corrosion process-reflecting physical quantity and the time series data of the first microbial community information using a machine learning model having the corrosion process-reflecting physical quantity as a target variable and the first microbial community information as an explanatory variable, and derives attribute information of the microbial community that may affect the corrosion of the metal material based on the selection result.

8. the machine learning model includes a regression equation, The corrosion factor detection system of claim 7, wherein the third derivation means derives a regression coefficient of the regression equation based on the time series data of the corrosion process-reflecting physical quantity and the time series data of the first microbial community information, selects the first microbial community information corresponding to the corrosion process-reflecting physical quantity based on the derived result, and derives attribute information of a microbial community that may affect the corrosion of the metal material based on the selected result.

9. The corrosion factor detection system according to claim 8, wherein the third derivation means derives the regression coefficients of the regression equation by performing a penalized regression analysis that can reduce the regression coefficients to 0 based on the time series data of the corrosion process-reflecting physical quantity and the time series data of the first microbial community information.

10. The corrosion factor detection system according to claim 7, wherein the third derivation means creates a sample by resampling using the time series data of the physical quantity reflecting the corrosion process and the time series data of the first microbial community information, and derives attribute information of the microbial community that may affect the corrosion of the metal material using the created sample.

11. the first microbial community information is information that can identify the amount of the microbial community for each type of the microbial community, 3. The corrosion factor detection system according to claim 1, wherein the attribute information of the microorganisms that may affect the corrosion of the metal material is information that can identify the type of the microorganisms that may affect the corrosion of the metal material.

12. the first microbial community information is information that can identify the function of the microbial community, The corrosion factor detection system according to claim 1 or 2, wherein the attribute information of the microbial group that may affect the corrosion of the metal material is information that can identify a function among the functions of the microbial group that may affect the corrosion of the metal material.

13. The corrosion factor detection system according to claim 1 or 2, wherein the microbial population includes microbial species derived from ribosomal RNA genes.

14. a first derivation step of deriving a corrosion process-reflecting physical quantity, which is a physical quantity whose value changes depending on the corrosion process of the metal material, by measuring a sample including a medium containing a microorganism group and a metal material in contact with the medium; a second derivation step of deriving first microbial community information, which is information that can identify at least one of the quantity and quality of the microbial community contained in the medium; a third derivation step of deriving attribute information of a microbial community that may affect the corrosion of the metal material based on the time series data of the corrosion process-reflecting physical quantity derived in the first derivation step and the time series data of the first microbial community information derived in the second derivation step; A corrosion factor detection method comprising:

15. A program for causing a computer to function as each of the means of the corrosion factor detection system according to claim 1 or 2.

Images