Food Contaminant Assessment

Abstract

The present invention relates to a computer-implemented method for predicting a risk associated with the presence of a contaminant in a food. The method can include receiving a query defining a food, a geography, and a contaminant. The method can then determine a risk that the food contains the contaminant relative to the geography. The method also includes providing a response to the query that includes one or more recommended management actions relative to the determined risk.

Description

Technical field and technical background

[0001] The present disclosure relates to methods and systems for the assessment of contaminants in food, and in particular, but not exclusively, to methods and computer-based systems for assessing the risk that contaminants are linked to food-related problems and suggesting appropriate control actions in relation to such contaminants.

[0002] The presence of contaminants such as pesticides or veterinary drugs in processed foods is a well-known risk in the food industry.

[0003] In this disclosure, the term "food" should be understood as corresponding to any material or materials intended for human consumption, whether processed, semi-processed or unprocessed, according to the definition given by Codex Alimentarius (2019 Codex Alimentarius Commission - Procedural Manual, 27th Edition, FAO and WHO, Rome, ISBN 978-92-5-131099-1): "Food" means any substance intended for human consumption, whether processed, semi-processed or unprocessed. It includes beverages, chewing gum and any substance used in the manufacture, preparation or treatment of "food", but does not include cosmetics, tobacco or substances used exclusively as medicines."

[0004] Examples of foods include: a) Foods of plant origin, including but not limited to raw or processed fruits, vegetables, herbs, and spices. Examples of foods of plant origin include fresh basil, apple puree, food coloring E160b derived from annatto, virgin olive oil, natural smoke aroma, and lecithin. b) Foods of animal origin, including but not limited to meat, eggs, dairy products, honey, whey powder, tuna, and egg yolks. c) Food from minerals such as, but not limited to, calcium carbonate, sodium chloride salts. d) Foods that are chemically, biologically or produced by modern technology, including but not limited to food colouring agent E133, artificial flavours, enzymes, lactic acid bacteria, baking soda, modified starches with enhanced gelling properties, genetically modified salmon.

[0005] Thus, "food" as discussed herein includes both standalone foods (whether single-component foods or foods prepared from multiple ingredients) and ingredients that are intended to be combined to produce a multiple-component food, such as in accordance with a food preparation recipe.

[0006] In the present disclosure, "feed" should be understood as corresponding to a product, single or multiple ingredients, whether processed, semi-processed or unprocessed, intended to be fed directly to animals, according to the definition in the Codex Alimentarius Standard (CXC 54-2004, Good Animal Husbandry Practice). It will therefore be understood that feed can be treated as equivalent to food, in the sense that all techniques and / or constraints described herein as applicable to food may also be applied to feed.

[0007] In the present disclosure, the term "contaminant" should be understood as corresponding to a substance that may be found in food or feed that may be deemed undesirable by one or more persons or one or more authorities. Thus, as used herein, "contaminant" includes: (i) Any substance that has not been intentionally added to food or to feed for food-producing animals and that is present in such food or feed as a result of the production (including operations carried out in crop management, animal husbandry and veterinary medicine), manufacture, processing, preparation, treatment, packing, packaging, transport or storage of such food or feed, or as a result of environmental contamination. In this context, the Codex-General Standard for Contaminants and Toxins in Food and Feed may also be relevant, which states: "Contaminant may mean any substance that has not been intentionally added to food and that is present in such food as a result of the production (including operations carried out in crop management, animal husbandry and veterinary medicine), manufacture, processing, preparation, treatment, packing, packaging, transport or storage of such food, or as a result of environmental contamination. The term does not include insect debris, rodent hair, and other foreign matter." (ii) Any substance intentionally and legally added to food or feed that may still cause food-related problems due to failure to follow recommended practices or due to incidents occurring during food production (such as, but not limited to, over-supplementation of an approved vitamin), or due to inconsistencies in the regulatory frameworks between the country of production and the country of sale (such as, but not limited to, food additives that are approved in one country but prohibited in the country where the foods containing them are ultimately sold). (iii) Any substance intentionally but unlawfully added to food or feed, including, but not limited to, adulterants.

[0008] Examples of contaminants include: a) Substances used to support plant- and animal-derived food production, or derived from such use, including, but not limited to, plant protection products, fertilizers, veterinary drugs and their metabolic and decomposition products. b) Substances that occur naturally in the environment, such as, but not limited to, mycotoxins, plant toxins, and chemical elements (such as, but not limited to, heavy metals). c) Substances occurring in the environment as a result of anthropogenic activities or technological catastrophes, including, but not limited to, persistent organic pollutants, pollutants resulting from incineration, and radionuclides. d) - Any contaminated surfaces or production equipment during production, transportation and storage, including but not limited to biocides, mold contamination of pallets, lubricants, etc. -Used packaging, including but not limited to mineral oil from gunny bags, plasticizers, semicarbazide in gaskets, etc. or by fumigation procedures, such as, but not limited to, food sterilization with ethylene oxide; Substances that enter the food or feed chain through direct or indirect contact. e) Substances produced during food or feed processing, in particular during heat treatment, frying, drying or fermentation, such as, but not limited to, Maillard reaction products, acrylamides, carbamates, furans, nitrosamines, chloropropanediol and its fatty acid esters, glycidol and its fatty acid esters. f) Substances added legally or illegally to food that may cause food-related problems, including but not limited to food additives, food processing aids and food adulterants. g) Microorganisms, such as, but not limited to, viruses and bacteria, which may enter the food chain at any step or result from not following the respective recommended practices during food production. h) Traces of food originating from specific technologies such as, but not limited to, biotechnology or nanotechnology, such as, but not limited to, genetically modified crops, titanium dioxide used in nanoparticle form, which differ from their conventional counterparts.

[0009] Thus, such contaminants may originate from a wide variety of sources along the food preparation chain. Indeed, throughout the food preparation chain, a food may become contaminated with multiple contaminants. Furthermore, when multiple foods are used as ingredients for a multi-ingredient product, each of those ingredients may contain multiple such contaminants. Thus, both single-ingredient foods and foods prepared from multiple ingredients may ultimately contain amounts of contaminants that exceed those permitted in the market in which the food is sold, or that otherwise exceed safe amounts, or that otherwise cause food-related problems.

[0010] As will be appreciated, foods (whether single-ingredient foods or foods prepared from multiple ingredients) are subject to rigorous quality assessments by both manufacturers and independent laboratories. Such assessments aim to ensure that the food meets suitable food quality levels. It will be appreciated that food quality may relate to any one or more of food safety, food taste, food texture, food compliance, etc. If the presence of a contaminant in a food may cause a decrease in any such food quality measure, this is described as a food-related issue. Thus, such quality assessments may aim to ensure that such contaminants do not affect the quality of the food.

[0011] Despite the use of food quality assessments throughout the food production chain, there is still a risk that contaminants may remain at higher than desired levels when the food reaches the consumer. This may be due, for example, to changes in regulations in a given country, new scientific discoveries that reveal new risks associated with compounds previously considered safe, etc.

[0012] Therefore, maintaining food quality (including, for example, safety) proves to be a significant challenge for the food industry. The operation of complex food supply chain channels leads to complex food quality solutions to minimize risks (for both companies involved in the food industry and food consumers) and ultimately to avoid product recalls, which currently occur at a rate of thousands per year worldwide. In the European Union alone, it is known that more than 3,400 food recalls occurred in 2020. The direct consequences of these recalls include food waste and associated costs that affect both the food industry and consumers.

[0013] Several methods and systems have been developed that provide a general ranking of pesticides (such as the "Checkyourscope" system provided by the EU Reference Laboratories, see https: / / www.eurl-pesticides.eu / docs / public / tmplt_article.asp?LabID=200&CntID=746&Lang=EN), allowing laboratories to prioritize them for inclusion in their targeted analytical portfolios. However, this excel-based system, based on the assignment of points based on toxicology, residue status in crops and agricultural use, only provides a recommendation for the inclusion of a given pesticide in the laboratory's analytical portfolio. Overview

[0014] The present invention relates to a computer implemented method, system and associated readable medium for predicting the risk of a contaminant being present in food or feed, as defined in claim 1 and below.

[0015] In particular, the present invention provides a computer-implemented method for predicting and managing risks associated with the presence of a contaminant in a food or feed product, comprising: i) providing at least one ingredient present in a food or feed product from a user query and at least one country of sale of the product; ii) obtaining the geographic origin of at least one ingredient; and iii) determining a risk score that the food or feed contains a contaminant in relation to the ingredients present in the food or feed product and their respective geographical origin, wherein determining at least one query data selected from the list consisting of a contaminant type, a food or feed type, and a geography type; at least one input data selected from the list consisting of contaminant specific relevance, likelihood that the contaminant will be used or found in the food environment, contaminant maximum residue level, contaminant maximum limit status, analytical data, and contaminant origin characteristic; determining whether the combination of iv) providing a response to the query that includes one or more specified management actions related to the determined risk; If the score corresponds to no risk being identified, An administrative action corresponds to the authorization to incorporate an ingredient into a food or feed product, or If the score corresponds to an identified risk that requires the ingredient not to be used, The management actions are a) sourcing ingredients from different suppliers; b) conducting further detection tests to verify whether the identified contaminant is present in a particular batch of ingredient; c) modifying the ingredients and / or the production process of the food or feed product to eliminate the identified contaminants; d) removing the ingredient from the food or feed ingredient list; corresponds to a recommendation to take at least one of the following actions: To provide and The present invention relates to a method comprising the steps of:

[0016] The disclosed approach provides methods and systems that can predict the risks associated with a wide variety of contaminants in food and / or feed and suggest appropriate management actions in relation to such contaminants. Such an approach can take into account various parameters that can affect the quality assessment of a given food or feed.

[0017] Thus, the approach provides an integrated food / feed quality system solution that allows linking the in silico assessment of the risk of contaminant occurrence to food / feed related problems and outputting suggestions for appropriate management actions. Thus, the in silico analysis for the presence / risk of contaminants can be performed before or during steps of the food / feed production chain such as the commissioning of the food / feed production, the production of the food / feed, the harvesting of the food / feed, or the transportation of the food / feed, so that any risk management / mitigation actions that need to be adopted can be taken before the food / feed is prepared for sale, and possibly before the food / feed is transported and / or grown.

[0018] Viewed from a first aspect, there can be provided a method for predicting a risk associated with the presence of a contaminant in a food or feed product, the method comprising receiving a query defining a food or feed product, a geography, and the contaminant, determining a risk that the food or feed product contains the contaminant in relation to the geography, and providing a response to the query including one or more specified control actions in relation to the determined risk.

[0019] Viewed from another perspective, there can be provided a system for predicting a risk associated with the presence of a contaminant in a food or feed product, comprising a programmable computer unit configured to receive a query defining a food or feed product, a geography and a contaminant, determine a risk that the food or feed product contains the contaminant in relation to the geography, and provide a response to the query including one or more specified control actions in relation to the determined risk.

[0020] The present teachings may be provided by a computer-readable medium carrying instructions for performing such methods and / or for programming a computer as such a system. [Brief description of the drawings]

[0021] Exemplary implementations of the approaches taught herein are described with reference to the following drawings. [Figure 1] FIG. 1 is a schematic diagram of an exemplary system for deploying the present approach. [Diagram 2] FIG. 2 is a schematic diagram of a data definition structure for a query dimension data definition that may be used with the present approach. [Diagram 3] FIG. 3 is a schematic diagram of the process flow of this approach. [Figure 4] FIG. 4 is a schematic diagram of a high-level decision tree that may be used in the present approach. [Diagram 5] FIG. 5 is a schematic diagram of a data definition structure for an output action definition that may be used with the present approach. [Figure 6] FIG. 6 is a schematic diagram of a management action suggestion decision chart.

[0022] While the approach described herein is susceptible to various modifications and alternative forms, specific embodiments have been shown by way of example in the drawings and are described in detail herein. It should be understood, however, that the drawings and Detailed Description are not intended to limit the scope to the particular forms disclosed, but rather, the scope is to encompass all modifications, equivalents, and alternatives falling within the spirit and scope defined by the appended claims. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0023] The present approach relates to analyzing one or more large-scale, non-homogenous datasets against a query set of three or more dimensions. To facilitate such analysis for results that allow specific mitigation actions to be deployed, the datasets may be defined and characterized according to characteristics of the data, and the queries and / or datasets undergo content processing that allows the queries to be evaluated against relevant data from the datasets despite differences in nomenclature, format, and / or language of the data in the dataset compared to the query definition.

[0024] In the following discussion, the terms food and feed are used interchangeably and it will be understood that, unless otherwise stated or the context indicates otherwise, any reference to food may be replaced by a reference to feed.

[0025] More specifically, the present approach permits evaluation of queries that specify contaminants, foods of concern from which the contaminants may emerge, geographies from which the foods may originate, and / or geographies in which the foods are intended to be used / consumed / processed / sold (for brevity, all of these uses are referred to herein generally as being used and / or consumed). Such queries are evaluated against one or more large, heterogeneous datasets to determine the risk that the foods may contain such contaminants when sourced from and / or used / consumed in the specified geographies. From such determined risk, one or more response actions can be defined to permit management of the risk.

[0026] By utilizing the present approach, an in silico (computer-based) analysis can be achieved that can be used to mitigate or minimize the risk (which may additionally or alternatively be referred to as the likelihood) of food recalls and food crises by proactively maintaining a high level of quality and food safety compliance in products sold to consumers. By employing the present approach, traditional limitations of predicting the presence of contaminants that may cause recalls / crises can be addressed. The present approach can provide such results by meaningfully taking into account existing complex and heterogeneous data sets related to contaminants in food, such as various individual and overlapping regulatory frameworks across various countries / jurisdictions for the control of contaminants, and can also take into account potentially unknown food origins, thereby avoiding food quality issues that may arise from sparse or complete lack of data on which contaminants may result from the production of food and / or which contaminants may be present in food of a particular type and / or origin, and / or from limited resources available for food production oversight.

[0027] As shown in Fig. 1, a system 1 can be deployed to implement the present approach. The system is provided with a computer unit 2. In this example, this is a computer unit having a processor and a memory configured to execute instructions providing the functionality of the present approach.

[0028] The computer unit 2 has or is connected to a storage device 3. The storage device 3 in this example is a local storage device such as a magnetic hard disk or solid state drive, but in other examples can be any other data storage device, whether local to the computer unit 2 or remote storage such as network addressable storage, storage area network, or cloud storage. The storage device 3 can be used to store input query data, including queries regarding contaminants, foods, and geographies for which risk is determined, as described further below. In some examples, the storage device 3 may also store program instructions for use by the computer unit 2 in operation and / or store input data used in the evaluation of the query data.

[0029] The computer unit 2 is also connected to a network 4. To communicate with such network 4, the computer unit 2 may include one or more network communication interfaces. The network 4 provides access to remote data sources and / or inputs. Such remote data sources and / or inputs may include a source of a query specifying the query data. Such remote data sources may additionally or alternatively include a source of input data used to evaluate the query data. In this example, the network 4 provides connectivity to one or more data stores from which the input data may be accessed or obtained by the computer unit 2, and thus examples of the network 4 include LANs, WANs, the Internet, etc., through which the input data may be provided from the data stores. To obtain data from the remote data stores, the computer unit 2 may call a suitable API of the data store. The data retrieved from the data store may be part or all of an already created data set and / or may be a real-time or near real-time fetch of data from a dynamically updated data set. In other examples, the network includes the input data data store(s), and in such examples the network 4 may be a storage area network, a cloud data service, etc. In addition, it will be appreciated that input data received over the network may be stored or cached in the memory of the computer unit 2 and / or in the storage device 3 .

[0030] The computer unit 2 is also provided with one or more input devices 5 by means of which a user may provide input to the system. Examples include buttons, a keyboard, a mouse or other pointing device, etc. Other examples include ports into which physical data media may be inserted. Using the input device 5, a user may initiate a query evaluation. The query may already be stored in the storage device 3, may be entered directly using the input device 5, or may be received over the network 4.

[0031] The computer unit 2 is also provided with one or more output devices 6 by means of which the system may provide output to a user. Examples include displays, printers, audio output devices, etc. The output devices 6 are used by the system to provide feedback to the user regarding the progress of the system's execution, in particular the processing of queries. For example, the display may be used to implement a graphical user interface in which the user may provide input using the input device 5 and receive information from the display. An audio output device may vocally communicate the user interface and / or provide an audible alert, for example, to acknowledge an instruction and / or to indicate that the query process is completed. In some examples, the output of the query processing, including any management action suggestions, may be provided in a data file format that may be displayed via the output device, stored on a storage device, and / or transmitted over a network. Such data files may be formatted according to the software environment in which the user is intended to utilize the output of the query processing, such as a metadata format flat file (e.g., an XML file, such as used with systems such as SAP SCM software or SRM software), or a database readable file (such as used with database applications such as SAP HANA, Oracle Database, MySQL, etc.).

[0032] Additionally, the input device(s) 5 and / or output device(s) 6 may be provided by a mobile device having its own input and output devices that forward input received at the mobile device to the computer unit and forward output from the computer unit to the mobile device for display / output at the mobile device. Such a mobile device may communicate with the computer unit 2 via a wired or wireless connection, such as, for example, PAN, LAN, WAN and / or access network technologies.

[0033] Thus, in use, the system may receive, via the input device, a query to be resolved or a selection of an existing query, evaluate the query using the input data, and output risks and / or identified actions to be taken in relation to such risks. The output may be provided in whole or in part via the output device or may be communicated to another system using the network 4. The query data and input data may be variously provided from the storage device 3, the network 4 and / or the input device 5, and therefore the query data and input data are shown in Figure 1 with dashed lines leading directly to the computer unit 2.

[0034] Thus, as will be apparent, in this example, system 1 is implemented in a single networked computer system, in other examples the system may be implemented in a distributed computing environment with any user interface capabilities provided by some form of terminal device having a local interface that interfaces with systems in the distributed computing environment or accesses a web interface provided by one or more elements of the systems in the distributed computing environment.

[0035] It will thus be appreciated that system 1 includes processing capabilities that are capable of receiving a query (e.g., from direct user input, from local storage, from remote storage, or from a remote system), receiving input data usable to resolve the query (e.g., from direct user input, from local storage, from remote storage, or from a remote system), and processing the query to provide an output that includes one or more of a risk that a certain contaminant is present and identified actions to be taken in relation to such risk. The queries, input data, risks, and identified actions are discussed further below.

[0036] As Figure 1 shows, the query data in this example relates to contaminant types, food types, and geography types. Each of these elements or factors may be referred to as a dimension of the query.

[0037] Contaminants that the present approach can evaluate in a query may include any one or more of the contaminants defined above, including, for example, pesticides, veterinary drugs, toxic metals, biocides, mycotoxins, anthropogenic environmental contaminants, process contaminants, naturally occurring substances (such as minerals found in food-producing geographies), genetically modified organisms (GMOs), and the like. Also as noted above, contaminants may include derived terms, such as the definition of a metabolite or complex regulated residue of an active ingredient. Some further examples of contaminants that may be considered are aldicarb sulfone, Bacillus thuringiensis, 1-naphthylacetamide, and 1-naphthylacetic acid (represented as 1-naphthylacetic acid, the sum of 1-naphthylacetamide and 1-naphthylacetic acid and their salts). As explained above, all contaminants considered herein are substances that may be harmful, potentially harmful, or otherwise undesirable in food.

[0038] The types of food may include any one or more of the foods defined above, including foods used directly as single-ingredient foods, and / or foods used as ingredients of multi-ingredient foods, and / or multi-ingredient foods themselves. Further examples of foods may include plant materials (such as fruits, vegetables, spices, herbs, vegetable oils, etc.), animal materials (such as fish, meat, animal fats, etc.), or minerals (such as mined minerals, manufactured colors / flavors, etc.). Some examples of foods that may be considered are grains, corn, concentrated lemon juice with a Brix value of 32. As explained above, all types of foods considered herein are intended to form or be used as ingredients of foods.

[0039] Types of geographies considered herein include geographical and / or geopolitical divisions related to the production, use, and / or consumption of edible foods, beverages, and the like. Such geographical and / or geopolitical divisions may include continents, countries (e.g., nations), sub-regions of countries (e.g., states, sub-counties, counties, etc.), and / or food-producing regions of any relevant size (e.g., the Nile Delta, The Low Countries, wine-producing appellations, regions associated with particular Designations of Origin Controlled, etc.). Some examples of geographies that may be considered are the entire world, Southern Europe, and Brazil.

[0040] These query data are used to create specific queries regarding the risk that a particular contaminant (or contaminants) occurs in a given type of food that originates from and / or is used or consumed (e.g., sold) in a given geography. One exemplary query is the question, "Is it possible that aldicarb (contaminant dimension) associated with lemons (food dimension) originating from Brazil (geographic dimension) could lead to problems when sold in Germany (geographic dimension)?"

[0041] It should be noted that additional elements or factors (dimensions) such as food production schemes (e.g. organic, animal welfare standards, UTZ sustainable agriculture standards, Bio Suisse certification, etc.) may be included in the query. In addition, elements and / or factors (dimensions) such as those specific to a particular business or market may be taken into account. For example, compliance may vary depending on the field of use. For example, there may be different constraints for human adult nutrition products, human infant nutrition products, and human elderly nutrition products. Also, there may be different constraints for animal feed, for example, for production animals and for pets.

[0042] 2, a schematic data definition structure for query dimension data definition that may be used with the present approach is shown. As shown, the present embodiment considers the highest level inputs required to define a query, which are to define contaminant type, food type, and geography type.

[0043] As shown, the definition may optionally include one or more further definition levels for the query. For example, in the type of contaminant, the query may specify whether this is a pesticide, a veterinary drug, a naturally occurring substance. In the example of a naturally occurring substance, this may be specified as a heavy metal or a microorganism, etc. Some of these further definitions may be provided by the user entering the query data, for example, the user may specify whether the geography shown should be considered in the query as a food producing geography or a food user / consumption geography. Some of the further definitions may be known to the system, for example, if the user specifies the contaminant "cadmium", the system may populate various known levels specifying that this contaminant is "naturally occurring" and belongs to "heavy metals".

[0044] As Figure 1 also shows, the input data in this example relate to contaminant specific associations, risk of the contaminant being used or found in the food environment, contaminant maximum residue levels / limit status, analytical data, and contaminant origin properties. These elements or factors may be referred to as "decision support elements". Each decision support element includes or depends on several properties that allow characterization of each decision support element, as shown below.

[0045] Contaminant-specific relevance as used herein refers to a known or measured risk that a substance defined as a contaminant is inherently associated with one or more foods. In some implementations, this can be an indication as to whether the substance defined is actually considered a contaminant in terms of the need to define actions to manage the contaminant. In other words, contaminant-specific relevance can be evaluated at the level of the contaminant dimension of the query, as it can be considered an inherent property of the contaminant itself. For example, lecithins resulting from the use of plant protection products can be considered not relevant based on the fact that they are substances that already occur naturally in certain foods, such as vegetable oils. In another example, the substance chlorpyrifos, a well-defined pesticide known to result in residues in treated plants, can be considered to have an inherent relevance as a contaminant.

[0046] In addition to being taken into account in relation to the contaminant dimension of the query, contaminant specific relevance may also be taken into account in the context of other of the query dimensions, either on top of or instead of. For example, the criteria set to define the specific relevance of a contaminant may be country-specific (or otherwise geography-specific) or may relate to another query dimension, such as business / industry. One example may be a microorganism used as a biopesticide that likely has no specific relevance in one food production geography (e.g., where its use is permitted and no monitoring is required), but in another geography (e.g., where food consumption is intended) the same microorganism may be considered a risk and therefore may have specific relevance as a contaminant. In an implementation where GMOs are considered contaminants, the relevance of this geography dimension may be very high, since regulations on GMOs are an example of a potential contaminant that is currently very country-specific.

[0047] Another example related to the food dimension is that a contaminant may be considered to be inherently associated with the production of baked goods utilizing cereal grains, but the same contaminant is not considered to be inherently associated with the production of food in the form of distilled alcohol, since it is destroyed, removed or denatured by the alcohol distillation process. A further such example is that by heat treating food, the risk of microbiological contaminants may be eliminated and / or some antinutrient complex contaminants may be degraded. A more specific example is that the presence of acrylamide (a contaminant) in coffee can be reduced by treating coffee with certain enzymes.

[0048] Thus, contaminant-specific associations may be characterized by properties such as, for example, the food and / or geography to which the association applies, age / freshness characteristics of the association data, and the like.

[0049] As used herein, the risk of a contaminant occurrence being used or found in the food environment refers to the risk of having a given contaminant associated with a food and / or geographic origin as a result of the production (including operations performed in crop management, animal management and veterinary medicine), manufacturing, processing, preparation, treatment, packing, packaging, transportation, or storage of such food, or as a result of environmental contamination. This occurrence risk information can be provided from sources such as registrations of the use of a particular contaminant in food production in a given geography, and / or from scientific literature or studies on the presence of a substance in a given location, and / or from observations / discussions related to the field. For example, acetamiprid, which is registered in Switzerland for use on apples, can be used as a marker that this contaminant may be used or found in the food environment. In another example, scientific publications reporting on naturally occurring cadmium in Honduras can be used as a marker of the risk that this contaminant may be used or found in the food environment. As will be appreciated, this decision support element may be relevant to all of the food, geography, and contaminant dimensions of a query, although in some cases perhaps it is more tied to food geography of origin than to food use or consumption geography. Thus, the risk of a contaminant occurrence being used or found in the food environment may be characterized by various characteristics such as the food dimension, the geography dimension, the nature of the data source, the age of the information, the amount of relevant data, the diversity of the origin data, etc.

[0050] Contaminant maximum residue level / maximum limit status as used herein refers to regulatory information regarding maximum allowable limits and / or residue levels of a contaminant. This decision support element is applicable to the geography dimension in that such regulatory information may be set at a geopolitical level (such as by country). This may be relevant to food production geographies, since regulations regarding the presence of a particular contaminant may be a marker that this contaminant is used in a food production environment. This may additionally or alternatively be relevant to food use / consumption geographies, since regulations regarding acceptable residues / limits in food for consumption may be envisaged to be applied in relation to the sale / supply of food within such geographies. In addition, this decision support element may be applicable to the food dimension, since the regulatory information may be related to a particular food in which the contaminant is considered to be of concern from a regulatory perspective. However, the present approach may also take this decision support element into account without being related to the food dimension or in relation to a generalized / abstracted definition of the food dimension. For example, if a regulation exists for residues in pears, the decision support element may be taken into account for a broader range of foods, such as pears and apples, all woody fruits, all berries, all plant foods, etc. For example, a regulation stipulating that carbaryl is regulated in Canada at a limit of 7 mg / kg may be taken as a marker that carbaryl may occur in blueberries, with an additional consideration that may be considered is that this possible occurrence may apply to blueberries sold in Canada, blueberries grown in Canada, shrub fruits including blueberries, and / or all fruits. As will be appreciated, the absence of a regulatory limit may be considered as specific data available for this decision support element, since the absence of a regulatory limit may indicate a low risk of the presence of the contaminant and / or a low relevance of the contaminant. Similarly, corrective actions specified for a regulatory violation may indicate the significance of the contaminant as a contaminant. Thus, contaminant maximum residue levels / limits may be characterized based on characteristics such as the geography to which the regulatory limit applies, the food to which the regulatory limit applies, the regulatory penalties required if the limit is violated, the regulatory corrective actions required if the limit is violated, etc.

[0051] Analytical data as used herein refers to the presence (or absence) and content of any analytical data for one or more of the query dimensions, such as the amount of data, the number of observations of a particular result in the data, any restrictions on the data (such as quantification restrictions) and associated statistics, as well as any associated attributes / metadata. As will be understood, the absence of any analytical data may be considered as specific data available for this decision support element. An example may include data that there is no analytical data available for carbaryl in blueberries from Canada. 1000 data are available for glyphosate, 30 data collected over 5 years for traditional fresh peppers from Turkey are above 0.005 mg / kg for chlorpyrifos, etc. As these examples show, analytical data may be related to contaminants and / or related to other dimensions in addition to the contaminant dimension. As an example, analytical data may be characterized by several characteristics such as the name of the analyte, the food tested, the concentration, the year of analysis, and associated statistics. Thus, each of these characterizations can be taken into account when using analytical data in evaluating a query, for example specifying that only data from the last n years should be used, or specifying that an analyte must be detectable at a certain minimum concentration to be relevant.

[0052] Contaminant origin signature as used herein refers to the assignment of a contaminant to a particular list set by an external or internal body originating from any or all of the other decision support elements. Again, the absence of a contaminant origin signature may be considered as a particular data available for this decision support element. For example, if a given pesticide or other plant protection product is not included in any contaminant origin signature related to lethal, acute or long-term toxicity in a given geography (e.g., the geopolitical area covered by such internal or external body), the risk of that contaminant causing a problem if found in food may be considered to be essentially low (or lower than pesticides known or suspected to have lethal, acute or long-term toxicity, such as pesticides identified as possible carcinogens). For example, a list may define alicarb as an unacceptable pesticide according to the 4C certification standard for coffee. There are many other such food standardization organizations, operating in different geographic and / or food scopes. It will be understood that this decision support element is therefore related to the contaminant dimension and may or may not also be related to the food dimension and the geography dimension. Thus, contaminant-derived characteristics may be characterized by characteristics such as the awareness level of the organization that listed the contaminant, the food in which the contaminant was listed, or the geography in which the listing organization operates.

[0053] Illustrative examples of specific characteristics that may be used to characterize one or more of the decision support elements are provided in Table 1 below. [Table 1]

[0054] It will thus be appreciated that the present approach may be implemented to evaluate queries (in more than two dimensions) against the potentially vast amounts of heterogeneous data (e.g., millions of data points) that make up the decision support elements to address critical matters. Each of the decision support elements may be of varying relevance to different queries, and each individual query should be evaluated as completely and accurately as possible against all of the available data for each of the decision support elements. Also, as further described below, in order to provide such a complete and accurate evaluation of a query, the present approach may need to take into account the decision support element data in describing the same or similar entities in each of the query dimensions using a variety of different data formats, different languages, and different names or terms. Characterization of each of the decision support elements may also aid in the "big data" nature of the query evaluation process, as the characterizations may be defined to enable very different types of data within the decision support elements to be comparable.

[0055] Thus, the present approach permits evaluation of queries that specify contaminants, foods of concern from which the contaminants may emerge, geographies from which the foods may originate, and / or geographies in which the foods are intended to be used / consumed / processed / sold (for brevity, all of these uses are referred to herein generally as being used and / or consumed). Such queries are evaluated against one or more large, heterogeneous datasets to determine the risk that the foods may contain such contaminants when sourced from and / or used / consumed in the specified geographies. From such determined risk, one or more response actions can be defined to permit management of the risk.

[0056] 3 illustrates how the present approach uses decision support elements to evaluate queries and provide management suggestions. All steps shown in dashed lines may be considered optional in at least some implementations.

[0057] As shown in Figure 3, the method of the present approach uses these query data (identified in step S3-1) and, if necessary, prepares them for processing in step S3-3. Examples of pre-preparation include data format conversion, naming processing (which may include synonym processing and / or closely related data processing), and linguistic processing.

[0058] Data format conversion is used when a query is provided in a format that is not directly compatible with the processing steps used to evaluate the query. One example is when a query expresses quantitative parameters in imperial units, but metric / SI units are required for query processing.

[0059] Synonym processing may involve mapping parameters specified in the query data (such as names of pollutants, foods, and / or geographies) to possible synonyms of that data that may be used in the input data used to evaluate the query. For example, the food specified in the query data may be "courgette" and the synonym processing may map this to "zucchini" or any other term used to describe this food. In another example, the geography specified in the query data may be "The Low Countries" and the synonym processing may map this to "Belgium," "Netherlands," and "Luxembourg."

[0060] Closely related data processing may include expanding a definition in the query data to include items that are closely related to the specified definition. For example, in a contaminant definition, the closely related data processing may add other closely related compounds such as salt, ester, metabolite, decomposition product, complex residue definitions for that contaminant. All of these additional contaminant definitions may be additionally defined in the query, or the query may be adapted to an inclusive level that includes all such closely related compounds in the query evaluation process. Such an inclusive level may be referred to as a common denominator. One particular example is that the contaminants defined in the query may be "glyphosate sodium salt" and "glyphosate potassium salt," whereas in the input data that is to be used to evaluate the query, these may be analytical data related to "glyphosate" and "aminomethylphosphonic acid," and the input data for the contaminant maximum residue level / maximum limit status may specify "N-(phosphonomethyl)glycine, including metabolites aminomethylphosphonic acid, N[(acetylamino)methyl]phosphonic acid, and N-acetyl-N-(phosphonomethyl)glycine." In this example, the common denominator could be "glyphosate."

[0061] Closely related data processing may also be performed on food and / or geography definitions in the query data. For example, if food is defined as "blood orange," the common term "orange" or "citrus fruit" may be set to include other orange varieties such as "Seville orange" and "Satsuma," and if the common term is set as "citrus fruit," it may also be set to include "lemon." Similarly, if the geography is set to an area, such as a specific, relatively small geographic area, the common term may be expanded to include surrounding areas. For example, if the defined geography is "Norfolk, United Kingdom," the common term may be set as "East Anglia, United Kingdom" to include nearby geographies with similar growing conditions.

[0062] Linguistic processing may include expanding the definition in the query data to include the same definition in multiple languages. Linguistic processing may be linked to the geography specified in the query definition or may be applied more universally. For example, if the food of interest is grapefruit and the query is made by a German speaker, the query may be assumed to define "Grapefruit" as a food dimension. In this example, if the specified geography is for fruits imported from South America to Switzerland, the definition may be expanded to include translating this defined food into languages ​​related to the food source geography (such as Spanish (pomelo) and Portuguese (Toranja)) and / or to include translating this defined food into languages ​​related to the food use / consumption geography (such as French (pamplemousse) and Italian (popelmo)). Further translations may include other languages ​​spoken in the source geography (continuing the same example, these may include "Quechua" and "Gaurani"). Additionally or alternatively, translations may be provided for the term(s) used to define the contaminant dimension and / or the term(s) used to define the geography dimension. As well as translations specific to the geography specified in the query definition, the translations may also include languages ​​associated with a large volume of scientific publications, and may extend to include, for example, all languages ​​known to be used in the definition of food regulations in any country. The languages ​​may also or alternatively be limited by the scope of the datasets constituting the decision support element against which the query is evaluated; for example, if all data are in English, French, German, Spanish, Italian, Japanese, and Chinese, there may be no need to translate beyond these languages, even if the geography involved may imply further languages.

[0063] It should be noted that any or all of these data preparation methods described in relation to step S3-3 may alternatively or additionally be performed on input data constituting the decision support data used to evaluate the query and / or may be performed as part of the query evaluation process. There may be processes performed on input data of data extraction from various sources (databases, data files (e.g., Excel), reports (e.g., PDF), etc.) to process the input data into decision support data by adapting the input data to a standardized language by defined transformations and / or mappings to enable the various input data to serve as homogenized decision support data. In addition, it is anticipated that at least some of these data preparation methods (whether applied to the query, the decision support element data, or both) may be performed using trained AI approaches to efficiently optimize large amounts of data.

[0064] By performing such data preparation prior to evaluating the query, the method can substantially ease the difficulty of comparing the query definition against the very large amount of heterogeneous data that constitutes the decision support element. In other words, without such a data preparation step, a specific implementation of the method may have to be restricted to a much more limited set of data that constitutes the decision support element, and / or the data that constitutes the decision support element may require data curation or indexing to make the data more homogenous prior to query evaluation. The approach does not preclude the use of limited or curated / indexed / homogenized data sets, but rather, in such instances, the data preparation effort is replaced by the effort required to preprocess the data prior to query evaluation. In some implementations, it may be appropriate to implement a mix of the data preparation step S3-3 and a separate preprocessing of the data that constitutes the decision support element.

[0065] Thus, the present approach permits evaluation of queries that specify contaminants, foods of concern from which the contaminants may emerge, geographies from which the foods may originate, and / or geographies in which the foods are intended to be used / consumed / processed / sold (for brevity, all of these uses are referred to herein generally as being used and / or consumed). Such queries are evaluated against one or more large, heterogeneous datasets to determine the risk that the foods may contain such contaminants when sourced from and / or used / consumed in the specified geographies. From such determined risk, one or more response actions can be defined to permit management of the risk.

[0066] Referring again to FIG. 3, in step S3-5, the query is evaluated by a query evaluation process. To process the query, the query data is compared to and evaluated against the input data. The exact algorithm used for any given query will vary according to the nature of the query and the available decision support element data. For example, if a contaminant has high inherent relevance and there is a very high risk that the contaminant will be used or found in the food environment of the production geography, other decision support elements may not be particularly influential and a high risk result may be returned after taking only those decision support elements into account. However, in such a situation, it may still be appropriate to consider other decision support elements, for example, the contaminant maximum residue level / maximum limit status, and / or the characteristics derived from the contaminant may provide useful information on what action should be proposed to manage that risk.

[0067] Thus, it can be seen that the approach to evaluating a query is a context-dependent decision space rather than a defined algorithm. One example is that if a contaminant specified in the query is not actually considered a contaminant in the geography considered in the query, the risk may be assessed as low without reference to other data. In another example, the process may start with a complete list of many or all possible contaminants and then eliminate some of them from consideration by comparing with the different dimensions specified in the query and the various decision support elements, which may start with contaminant-specific relevance. In a further example, the process may progressively create only content relevant to certain dimension values ​​from the various input data of the decision support elements. It can be seen that different approaches may be applied to different dimension values, for example, for a certain level 2 or level 3 contaminant (see, for example, FIG. 2), it may be appropriate to start with all contaminants and eliminate some of the contaminants as the data is reviewed, while for other definition types the opposite may be true. Once the query(s) are processed, the output indicates a level of risk for the specified query dimension that one or more of the contaminants specified in the contaminant dimension should be considered a potential problem. This risk may be expressed as a binary yes / no type of risk presence, or a more gradated risk such as low / medium / high, or even more gradated risk such as a numerical scale or percentage.

[0068] Thus, at a global level, the attribute values ​​are combined to derive the risk that the contaminant is linked to a food-related problem as well as potential management actions. This approach works by including the use of any closely related compounds and / or common denominators for the contaminants if defined.

[0069] As an example, if the pesticide dichlorvos is registered for use in New Zealand for kiwi cultivation and has been detected in kiwis from New Zealand, and these kiwis are about to be sold in Switzerland, where dichlorvos is not authorised and therefore the maximum residue level of dichlorvos for kiwis on the Swiss market is lower than the level set in New Zealand, one possible outcome of the algorithm could be that there is a high risk that dichlorvos present in kiwis from New Zealand and sold in Switzerland is linked to a food-related problem, and (with a view to management action recommendations as in subsequent steps of the method) the proposed management action could be that analysis of dichlorvos is mandatory as a food release criterion.

[0070] As an example of how erroneous processing in a query evaluation can lead to erroneous conclusions, consider the situation where the pesticide aldicarb is registered for use in Mexico for grapefruit cultivation. If only its metabolite aldicarb-sulfone is detected in grapefruit from Mexico, and these grapefruits are sold in Switzerland, where aldicarb is not approved and is regulated, the present approach using closely related compounds and / or intersection processing of the data used in the query evaluation correctly points out that there is a high risk that aldicarb (the sum of aldicarb, its sulfoxide and its sulfone, expressed as aldicarb) is found in such grapefruit and control action may be required. In contrast, consider a hypothetical scenario in which these data are considered alone. This may allow erroneous conclusions such as "aldicarb is used, but no residues are found, aldicarb is not regulated and therefore may not be associated with food-related problems" to be determined, since the pesticide itself is not found in grapefruit. The present approach avoids such inaccurate conclusions, despite the sheer volume and heterogeneity of the data, by taking a holistic view of the data, e.g., by considering regulations regarding source and destination geographies separately and by viewing contaminants in terms of closely related compounds and / or common denominators.

[0071] This processing of queries is also shown as a high-level decision tree in Figure 4. As shown in this figure, the query (including the three dimensions of contaminant type, food type, and geography type, as well as any optional further dimensions) is evaluated against decision criteria that correspond to various decision support element data.

[0072] By performing this evaluation of the query dimensions against the decision criteria corresponding to the decision support element data, the process provides one or more of a shortlist of contaminants (as specified in the query dimensions of one or more queries) that require specific management actions and management action proposals.

[0073] As further shown in FIG. 3, in step S3-7, evaluation of the queries can output a shortlist of contaminants requiring specific control actions.

[0074] This step may be used, for example, in situations where a query specifies multiple contaminants and / or when a batch of queries are submitted at the same or similar time, e.g., a query may specify "all pesticides" as a contaminant, e.g., in situations where a user has little knowledge about possible contaminants for a given food.

[0075] This step may also be used in situations, for example, where a query specifies a single contaminant, but the use of closely related compound processing and / or intersection processing effectively expands the query contaminants to include several contaminants.

[0076] Whatever the risk rating scheme used in a particular implementation, if multiple contaminants are identified that have risks indicating that some form of action or mitigation or management action may be required, these may be collated as a shortlist in step S3-7.

[0077] Referring to step S3-9, whether or not there is a shortlist created in step S3-7, the method proceeds to determine one or more suggested management actions that may be provided to the user to mitigate or eliminate the identified risks. In this approach, the management actions may be determined based on contaminant risk alone or in combination with some of the characteristics related to the query dimensions and / or decision support elements.

[0078] For example, if the risk result indicates a very high risk or a potentially fatal risk, then further characteristic information may not be considered to conclude that an instruction not to use the food specified in the query should be given as an administrative action (as shown in step S3-11).Similarly, if the risk indicated is zero or approximately zero, then characteristic information may not be used to indicate that no administrative action is required.

[0079] In other examples, the characteristic information may be taken into account in one or more specified control actions. For example, if the risk result is associated with a moderate risk due to the presence of a particular pesticide, a surveillance plan to monitor the presence of the pesticide in food may be an appropriate control action. Other possible control actions may include adapting the analytical portfolio (as shown in step S3-13), triggering an audit for a given source or supplier, etc.

[0080] Such exemplary outputs based on risk determined by the query evaluation process are further illustrated in FIG. 5, which shows a schematic data definition structure for output action definitions that may be used with the present approach. As shown, the definitions may take these data into account at various levels. For example, within a level 1 output of control action suggestions, these may be divided at a second level between completely blocking the release of the food (consistent with step S3-11) and various risk mitigation control actions that may be applied without completely blocking the food (consistent with step S3-13). As further illustrated by the additional spaces and connectors, these may be further defined at further levels depending on any particular implementation.

[0081] Further details of possible management actions and how they are determined from the risks determined by the query evaluation process are discussed below with reference to FIG.

[0082] In the final step shown in FIG. 3 (S3-15), the method includes validation feedback regarding successful execution of the control actions. Such validation feedback may be provided through computer-based tracking of the specified control actions and detectable indicators that they have been implemented. In some examples, this may take the form of data that may be included as further decision support element data for future evaluations, including data describing the behavior and / or impact of the control actions. In some examples, this validation feedback may alternatively or additionally be provided by specific data capture regarding the execution of the control actions. By including such validation feedback, the present approach allows the system to facilitate tracking of the implementation and impact of the control actions. For example, over the course of processing several queries and recommending certain control actions, it may be possible to learn that certain control actions have a higher success rate in preventing food recall events, etc., and thus the control action decisions may be modified over time to learn from these current historical results.

[0083] In some implementations, it may be appropriate to include some form of checking process in relation to the algorithmic results. For example, a new algorithmic approach is being tested and some form of double-checking of the results may be included as an algorithmic validation process. In further examples, such double-checking may be applied on an ongoing basis. Such checks may be applied directly at the algorithmic output (e.g., between steps S3-5 and S3-7) or in relation to proposed management actions and / or their validation (e.g., in or after steps S3-9 and / or S3-15).

[0084] One example would be to compare the algorithm output with results determined using a small but representative data set to check whether the results are entirely different from those expected from the small data set. This small data set may be a subset or may involve parallel or otherwise similar queries on different data. A particular example is a query being performed for contaminants in spices, which query relies on evaluation against a large amount of data. The check may be performed using just one spice as a representative subset, and the process may then be scripted and may optionally include the generation of intermediate data files so that checks can be performed at multiple different stages of data processing.

[0085] Another approach involves considering control actions and validation that occurs after the control actions are taken. The algorithm may be considered correct if validation indicates, for example, that expected contaminants have been detected via monitoring. Such an approach may be used "live" by performing ongoing monitoring of the algorithm's accuracy, but also in a "test" mode where the algorithm is used to evaluate queries against data where control action results are already available that can be used to validate the control action proposal.

[0086] Another possible approach is to use the output of the algorithm and / or the proposed management actions for a given query to predict regulatory changes relevant to the subject of the query, and evidence of the accuracy of the algorithm is then found in the regulatory changes. Again, this approach may be used "live" to look for possible future regulatory changes, or in a "test" mode where the fact of regulatory change is explicitly excluded from the decision support data and then run to see if the algorithm can predict regulatory changes.

[0087] Referring now to Figure 6, this provides a schematic diagram of a control action suggestion decision schema. This illustrates an approach that can be taken to utilize the determined risks for the query dimensions in combination with the data of the decision support elements used to process the query to select an appropriate control action. As will be appreciated, "control" actions herein do not refer to control in the sense of business control or the like, but rather to preventative or corrective actions and tasks taken to maintain food safety by treating food in a manner that minimizes consumer risk and the risk that a product recall or the like will be required.

[0088] The particular example schema shown in Figure 6 applies to plant protection products such as pesticides. Those skilled in the art will understand that these principles can be applied to any form of contaminant based on any given implementation of the present approach, based on the particular data at hand.

[0089] As shown in Figure 6, various shapes are plotted on top of each other to define regions of the decision space, with each shape corresponding to a property found in or characterizing one or more of the decision support elements. The regions defined by these overlapping shapes in the decision space are then used as boundaries to determine the appropriate management action to take depending on the values ​​of each of the properties that the shapes define.

[0090] In this example, the shapes represent the following types of data: [Table 2]

[0091] As shown, the size and relative placement of these shapes in the decision space becomes a schema for determining control actions. As noted above, the size of each shape, and therefore the area defined by their overlap, is not proportional to the number of cases that fall into the combinations created by the overlapping shapes. The different control actions identified by the various mappings of this exemplary schema are as follows: · Unacceptable situations, which require analysis as release criteria and corrective actions. Analyze for launch Monitor once a year Research and analysis is required If part of a screening method, analyze annually; otherwise, no action required Audit / supplier assurances that they are not used in the food chain Expert risk assessment No action required Low probability of occurrence, manual assessment (e.g. the combination of factors leading to a decision in this area is inconsistent or otherwise manual intervention is appropriate to double-check the results and, if necessary, debug the algorithmic approach).

[0092] Each of these mappings is shown in Figure 6 and will not be repeated in full here, however, by way of example, the control action applied when the product registration status is Not-Registered is "No action required" if the data point falls within the "outside" region of each of the Plant Protection Product Regulatory Status, Analytical Data Relationship Data Availability Threshold, Analytical Data Finding Threshold, and Characteristics of Origin, regardless of inherent relevance.

[0093] Also, by way of example, if the inherent relevance is Not-Relevant, all defined areas result in one of the following: "Expert Risk Assessment", "No Action Required", "Unlikely to Occur", or "Manual Assessment". Similarly, if the inherent relevance is Relevant and the originating characteristic is in the inner area, the management actions are "Unacceptable Situation", "Analysis Required as Release Criteria and Corrective Action" or "Monitor Annually".

[0094] As noted above, this schema is provided as an example, and in practice this schema may need to be determined for each implementation or group of implementations.

[0095] It will be appreciated that while the illustrated schema is visually represented in this figure, an actual plot need not be created to determine suitable management actions, rather the data may be analyzed to determine the outcomes that would have occurred if the actual data values ​​were plotted on a visual representation of the decision space according to the schema.

[0096] Some examples of specific exemplary situations and outcomes to which the present technology may be applied include the following: ● Finding glyphosate in lentils, even below the maximum residue limit, is still perceived negatively and could create food-related problems. A suggested control action could be to source raw materials from countries where glyphosate is not used and to audit the lentil producers. Verification of properly performed control action could consist in obtaining documentation providing full traceability of the lentils' origin and in passing an audit from the producer's side. ● Determination of the risk of occurrence of veterinary drug residues in dairy raw materials of different countries of origin, allowing the optimization of veterinary drug residue management in dairy supply to avoid food-related problems, as a proposal for control actions. This may include the need to vary the frequency of analytical testing on supplied dairy materials for a specific country of origin and, in certain cases, to investigate the use of veterinary drugs in primary production. Alternatively, this may include rejecting batches of raw materials that may result in food-related problems, corrective actions by the producer to improve the situation, or the exclusion of the primary producer from the raw material supplier. Validation may utilize test results (data) that improve the accuracy of the risk of occurrence of veterinary drug residues obtained at the proposed frequency of analytical testing, which is consequently included in the algorithm, or an audit report summarizing information on the practice of veterinary drug use in primary production. ● Definition of the risk of occurrence of lead and copper residues in honey from different countries of origin, leading to the proposal of control actions to avoid food-related problems. Such control actions may include optimizing the frequency of analytical testing for those honey contaminants depending on its country of origin, or rejecting raw material batches with levels of contaminants that may lead to food-related problems. The validation makes use of the test results (data) obtained at the proposed frequency of analytical testing and consequently includes them in the algorithm, which can improve the accuracy of the determination of the contaminant occurrence. ● Prediction / extrapolation of the risk of occurrence of veterinary drug residues and pesticides in beef products originating from a specific country where the product has not been supplied before. This may lead to the proposal of management actions to establish the frequency of control for the substances under investigation or to the rejection of raw material batches with levels of contaminants that may lead to food-related problems. Validation may involve utilizing the test results (data) obtained at the proposed frequency of analytical testing and consequently including them in the algorithm to improve the accuracy of the determination of contaminant occurrence.

[0097] Thus, various approaches have been described that can analyze large amounts of heterogeneous data to determine the risk that one or more types of contaminants are present in a type of food from and / or for a particular type of geography. These approaches allow the large amounts of data to be effectively used to determine one or more specific control actions to be used to improve food safety and / or reduce the risk that a food recall will be required, based on the particular data used to assess the risk and contaminants.

[0098] As will be appreciated, in that the present approach has been described in the context of computer implementation (which may be referred to as "in silico"), a computer program product may be provided that includes instructions executable to cause a programmable computer to perform the present method. Such instructions may correspond to the operating instructions of the computer unit 2 described above with reference to FIG. 1. Such a computer program product may be provided by a computer-readable medium. The computer-readable medium may be a storage medium and / or a transmission medium. The computer-readable storage medium may include magnetic storage media, optical storage media, and / or electronic data storage media, may be portable (CD, DVD, floppy disk, usb memory stick, memory card, etc.) or may be intended to be installed within a computer system (hard disk drive, SSD, etc.), and may generally be referred to as a "non-transitory" computer-readable medium. The transmission medium may be generated to carry instructions between components of a computer system (such as a bus, cable or other interconnect) and / or between multiple separate computer systems (such as via a wired or wireless network, an access network, a point-to-point cable, etc.), and may include a carrier wave, a transmission signal, etc.

[0099] In one aspect, a system and method for predicting a risk associated with the presence of a contaminant in a food has been described. The method can include receiving a query defining a food, a geography, and a contaminant. The method can then determine a risk that the food contains the contaminant relative to the geography. The method also includes providing a response to the query that includes one or more recommended management actions relative to the determined risk.

[0100] It should be understood that various changes and modifications to the illustrated examples for implementing the present approach described herein will be apparent to those skilled in the art. Such changes and modifications can be made without departing from the spirit and scope of the present invention and without diminishing its attendant advantages. Thus, such changes and modifications are intended to be encompassed by the appended claims.

[0101] Further embodiments and feature combinations of the present teachings are presented in the following numbered clauses. [Clause 1] 1. A method for predicting a risk associated with the presence of a contaminant in a food or feed product, comprising: receiving a query defining a food or feed, a geography, and a contaminant; determining the risk that said food or feed contains said contaminant in relation to said geography; providing a response to the query including one or more specified management actions associated with the determined risk; A method comprising: [Clause 2] 2. The method of clause 1, wherein said response further comprises said risk determined. [Clause 3] obtaining one or more characteristic data from a data store describing characteristics of at least one of the food or feed, the geography, and the contaminants; 3. The method of any one of clauses 1 to 2, wherein said determining comprises utilizing said one or more characteristic data to determine said risk. [Article 4] 4. The method of claim 3, wherein said characteristic data comprises data describing the risk that said contaminant may be found in food or feed originating from said geography. [Article 5] 5. The method of claim 3 or 4, wherein the characteristic data comprises data describing the inherent relevance of the contaminant as a problematic component of food or feed. [Article 6] 6. The method of any one of clauses 3 to 5, wherein the characteristic data comprises data describing the presence of analytical data associated with one or more of the food or feed, the geography and the contaminants. [Article 7] 7. The method of any one of clauses 3 to 6, wherein said characteristic data comprises data describing regulatory limits or thresholds for said contaminants in food or feed. [Article 8] 8. The method of any one of clauses 3 to 7, wherein the property data comprises a property originating from the contaminant. [Article 9] the characteristic data includes data from a plurality of different sources; 9. The method of any one of clauses 3-8, wherein the method compares data from a first source with data from a second source using a mapping scheme and / or a data term synonym protocol. [Article 10] 10. The method of any one of clauses 1 to 9, wherein the contaminant comprises a substance that may be contained within a food or feed, which substance is potentially harmful upon consumption of said food or feed by humans or animals. [Article 11] 11. The method of claim 10, wherein said determining includes considering said contaminant and compounds closely related to said contaminant. [Article 12] 12. The method according to any one of clauses 1 to 11, wherein the geography is the geographical and / or geopolitical area in which the food or feed is grown, harvested, processed and / or manufactured. [Article 13] 13. The method according to any one of clauses 1 to 12, wherein the food or feed is a substance intended for and / or used as an ingredient in an edible product. [Article 14] The recommended management action is: preventing the release of said food or feed; authorise the release of said food or feed; and / or implement mitigation actions in relation to the procurement or production of said food or feed; 14. The method of any one of clauses 1 to 13, comprising the instruction: [Article 15] After receiving the query, Data format conversion, Naming process, Common denominator processing, linguistic processing of one or more of said food or feed, said geography, and said contaminants; 15. The method of any one of clauses 1-14, further comprising performing one or more of the following: [Article 16] 16. The method of claim 15, wherein the data format conversion includes adapting the query to one or more data formats required for the determining. [Article 17] 17. The method of claim 15 or 16, wherein the naming process comprises one or more of a synonym process and / or a closely related data process. [Article 18] 18. The method of claim 17, wherein the synonym processing includes expanding the query to include synonyms of food or feed, the geography, and the contaminant. [Article 19] 19. The method of claim 17 or 18, wherein the closely related data processing comprises expanding the query to include definitions of compounds closely related to the contaminant, such as, for example, salts, esters, metabolites, degradation products, and complex residues for that contaminant. [Article 20] 20. The method of any one of clauses 15-19, wherein the language processing includes expanding the query to include a translation of at least one of the food or feed, the geography and the contaminants into another language. [Article 21] 3. The method of claim 3, or any one of the clauses subordinate thereto, wherein the characteristic data comprises heterogeneous data from a plurality of data sources. [Article 22] 3. The method of claim 3 or any one of the clauses subordinate thereto, wherein the characteristic data comprises at least 1 million data points. [Article 23] 23. The method of any one of clauses 1 to 22, further comprising outputting a shortlist of contaminants requiring specific management actions. [Article 24] 24. The method of any one of clauses 1-23, further comprising verifying performance of one or more of the one or more specified management actions. [Article 25] 25. A system for predicting a risk associated with the presence of a contaminant in food or feed, comprising a programmable computer unit configured to carry out the method according to any one of clauses 1 to 24. [Article 26] A computer readable medium carrying instructions which, when executed by a programmable computer, configure the programmable computer to cause the programmable computer to perform the method of any one of clauses 1 to 24. [Article 27] 1. A method for predicting a risk associated with the presence of a contaminant in a food or feed, said method comprising: (i) providing a set of input data from a local or network / internet database, said data including contaminant type associated with the raw material, geographic type, and food or feed type; (ii) assessing the risk associated with the presence of the contaminant in the food or feed in relation to the type of geography; (iii) providing a level of risk associated with the contaminant; (iv) providing a set of recommended actions associated with said risks; and A method comprising: [Article 28] 28. The method of claim 27, wherein the type of contaminant comprises a pesticide, a veterinary drug, a man-made environmental contaminant, or a molecule that occurs naturally in the environment. [Article 29] 29. The method of claim 28, wherein the pesticide comprises a chemical pesticide or a bio-based pesticide. [Article 30] 29. The method of claim 28, wherein the molecule is a man-made environmental pollutant such as a polyaromatic hydrocarbon or other incineration or recycling by-product. [Article 31] 29. The method of claim 28, wherein the molecules naturally occurring in the environment include heavy metals or microorganisms. [Article 32] 32. The method of any one of clauses 27 to 31, wherein the type of geography comprises crop producing countries and / or food selling countries. [Article 33] 33. The method of any one of clauses 27 to 32, wherein the food type comprises a fruit, a vegetable, or an animal ingredient. [Article 34] The set of recommended actions is: preventing the use of said raw materials associated with said contaminants; or The method of any one of clauses 27 to 33, comprising either adding the contaminant to an automated supervision plan, and / or adapting an analytical portfolio, and / or mitigating the risk by triggering an audit.

Claims

1. 1. A computer-implemented method for predicting and managing risks associated with the presence of a contaminant in a food or feed product, comprising: providing at least one ingredient present in said food or feed product from a user query and at least one country of sale of said product; obtaining a geographic origin of the at least one ingredient; determining a risk score for said food or feed containing a contaminant in relation to said ingredients present in said food or feed product and their respective geographical origin, said determining comprising: at least one query data selected from the list consisting of contaminant type, food or feed type, and geography type; at least one input data selected from the list consisting of contaminant specific relevance, likelihood that the contaminant is used or found in the food environment, maximum residue level of the contaminant, contaminant maximum limit status, analytical data, and contaminant origin characteristic; It is based on a combination of , determining providing a response to the query including one or more specified management actions associated with the determined risk score; If the risk score corresponds to no identified risk, the control action corresponds to an authorization for incorporation of the ingredient into the food or feed product; or If the risk score corresponds to an identified risk requiring the ingredient not to be used, the control action is: sourcing said ingredients from different suppliers; conducting further detection tests to verify whether the identified contaminants are present in the particular batch of ingredients; modifying the production process of said ingredient and / or food or feed product to eliminate the identified early-stage contaminants; removing said ingredient from said food or feed ingredient list; corresponding to a recommendation to take at least one of the following actions: To provide and 4. A computer-implemented method comprising:

2. obtaining one or more characteristic data from a data store describing characteristics of at least one of the food or feed, the geography, and the contaminants; The method of claim 1 , wherein said determining comprises utilizing said one or more characteristic data to determine said risk.

3. The method of claim 2 , wherein the characteristic data comprises data describing the inherent relevance of the contaminant as a problematic component of food or feed.

4. 4. The method of claim 2 or 3, wherein the characteristic data includes data describing the presence of analytical data associated with one or more of the food or feed, the geography, and the contaminants.

5. The method of any one of claims 2 to 4, wherein the property data comprises a property of origin of the contaminant.

6. 6. The method of claim 3, wherein the characteristic data comprises data from a plurality of different sources, and the method compares data from a first source with data from a second source using a mapping scheme and / or a data term synonym protocol.

7. contaminants include substances that may be contained in food or feed, said substances being potentially harmful upon consumption of said food or feed by humans or animals; For example, the determining step includes considering the contaminant and compounds closely related to the contaminant.

8. The recommended management action is: preventing the release of said food or feed; authorise the release of said food or feed; and / or implement mitigation actions in relation to the procurement or production of said food or feed; The method of any one of claims 1 to 7, comprising the instructions:

9. After receiving the query, Data format conversion, Naming process, Common denominator processing, The food or feed, linguistic processing of one or more of said geography and said contaminants; The method of any one of claims 1 to 8, further comprising performing one or more of the following:

10. The method of claim 9 , wherein the data format conversion comprises adapting the query to one or more data formats required for the determining.

11. The naming process includes: synonym processing, for example including expanding the query to include synonyms of the food or feed, the geography, and the contaminant; and / or Closely related data processing, including expanding the query to include definitions of compounds closely related to the contaminant, such as salts, esters, metabolites, degradation products, and complex residues for the contaminant. The method of claim 9 or 10, comprising one or more of the following:

12. 3. The method of claim 2 or any claim dependent thereon, wherein the characteristic data comprises heterogeneous data from multiple data sources and / or comprises at least 1 million data points.

13. The method of any one of claims 1 to 12, further comprising verifying the performance of one or more of said one or more specified management actions.

14. A system for predicting the risk associated with the presence of a contaminant in food or feed, comprising a programmable computer unit configured to carry out the method according to any one of claims 1 to 13.

15. A computer readable medium carrying instructions which, when executed by a programmable computer, configure the programmable computer to cause the programmable computer to perform the method of any one of claims 1 to 13.

Images