System and method for determining a protective element
Patent Information
- Authority / Receiving Office
- EP · EP
- Patent Type
- Applications
- Current Assignee / Owner
- HILTI AG
- Filing Date
- 2024-06-13
- Publication Date
- 2026-05-06
AI Technical Summary
Existing methods for selecting protective elements for building structures are time-consuming and reliant on expert training and experience, often resulting in inadequate or costly protection, especially in application situations not covered by certified elements.
A system comprising an input device, output device, and database that processes input data sets against comparison data sets to determine suitable protective elements, using image processing, natural language processing, and machine learning to assign or output identifiers for protective elements, ensuring reliable and cost-effective protection.
Enables efficient, reproducible, and quality-assured selection of protective elements, reducing the risk of incorrect installations and providing reliable protection for building structures, even in unique application situations, while being cost-effective.
Smart Images
Figure EP2024066334_02012025_PF_FP_ABST
Abstract
Description
[0001] System and method for determining a protective element
[0002] Description
[0003] The invention relates to a system for determining a protective element of a protective element type for use in an application situation relating to a building component.
[0004] To protect building components from various hazards, such as fire, various protective elements are available. These protective elements are each designed for specific application situations and are certified specifically for these situations.
[0005] However, when planning structural components, application situations may arise that do not directly correspond to any of these specific application situations and for which no specifically certified protective element is available. To ensure adequate protection for such an application situation, complex individual analyses are necessary. These can often be very time-consuming. The results, and thus the safety and service life of the structural components, can therefore depend significantly on the training and experience, as well as the concentration, of those conducting these individual analyses.
[0006] The object of the present invention is therefore to provide a system and a method that enable reliable protection of structural components. Systems and methods that also enable particularly cost-effective protection of structural components are particularly preferred.
[0007] The object is initially achieved by a system for determining a protective element of a protective element type for use in an application situation relating to a building part, comprising an input device, an output device and a database, wherein the database comprises a plurality of comparison data sets, each of which comprises data relating to a comparison application situation, wherein the system is configured to receive an input data set comprising data relating to the application situation via the input device and, depending on the input data set, to output at least one of the comparison data sets, a part of at least one of the comparison data sets, a reference to at least one of the comparison data sets and / or at least one designation of a protective element of the protective element type via the output device.
[0008] The input data set contains data from the application situation. The system can thus enable the assignment and / or output of one or more comparison data sets based on the input data set. Even if no identical comparison data set, and thus no protection element directly certified for the application situation, is available for the application situation represented by the input data set, a protection element or its identifier suitable for this application situation can be determined based on the one or more comparison data sets.
[0009] The system can also be configured to immediately output a designation for one or more suitable protective elements.
[0010] The system allows a large number of comparative data sets to be evaluated cost-effectively, reproducibly, and with quality assurance. The structural component can then be equipped with the protective elements determined in this way, ensuring that the structural component can be reliably protected by the appropriate protective element. By cost-effectively determining the protective element, this reliable protection can also be achieved particularly cost-effectively. Incorrect fitting of structural components with unsuitable protective elements due to lack of training or experience, fatigue, or similar factors is avoided.
[0011] The structural component may comprise part or all of a building or civil engineering structure, a steel structure such as an oil drilling platform, or the like. The structural component may, for example, be or comprise a wall, a ceiling, and / or a floor of the structure.
[0012] The comparison data sets may also include one or more results of the system from previous determinations of comparison data sets and / or appropriate safeguards for previous application situations. The comparison data sets may be selected to meet certain minimum quality requirements. For example, they may be limited to results of previous individual analyses by specific experts who, for example, meet certain educational requirements and / or have certain minimum experience.
[0013] The comparison data sets can preferably relate to application situations for which protective elements of the desired protective element type are available and / or certified. If such a comparison data set can thus be assigned to the input data set, a suitable protective element can be directly derived from the comparison data set. For this purpose, the database can contain at least one identifier of an assigned protective element of the protective element type for at least one, preferably for all, of the comparison data sets.
[0014] The system can be configured to assign one or more comparable comparison data sets to the input data set and / or output them to the output device. The comparison data sets and / or associated identifiers of protection elements can then be output to the output device.
[0015] If no comparable comparison data set is found, the system can also be set up not to assign a result or to output an error signal indicating that no comparable comparison data set is available and / or to output it on the output device.
[0016] When planning building components, a wide variety of application situations can arise. To describe these diverse application situations, image data, for example in the form of CAD drawings, photographs, sketches, or the like, can be available and / or created. Such image data can depict application situations in great detail and precision. For example, geometric relationships can be described simply and precisely using image data. Therefore, it is advantageous if the input data set and / or at least one of the comparison data sets includes image data.
[0017] The system can then preferably be configured to process image data of the input data set and / or one or more comparison data sets. For this purpose, the system can comprise an image processing unit and / or an image and speech processing unit, for example in
[0018] Form of a vision language model. Certifications of protective elements often take place for certain categorical properties of the application situations, for example for certain types of material, for certain positions such as installation in a wall, on a ceiling, or the like, for certain protection levels, for example for certain resistance classes, or the like. For this purpose, it can be provided that the input data set and / or at least one of the comparison data sets comprises nominally scaled data. The system can be configured to process such nominally scaled data of the input data set and / or at least one of the comparison data sets. Nominally scaled data also offers the advantage that it allows the input data set and / or the comparison data sets to be directly categorized.The input data set can thus be compared very easily and quickly, at least with regard to properties with nominally scaled data, with one or more of the reference data sets.
[0019] To enable the greatest possible flexibility of the system, it can also be provided that the input data set and / or at least one of the comparison data sets comprises text data. Accordingly, the system can be configured to process such free-text data. For this purpose, the system can comprise a natural language processing module, hereinafter referred to as an NLP module.
[0020] To select and / or classify comparison data sets, the system can be configured to determine a subset of the comparison data sets depending on the input data set. In particular, it can comprise a classifier. It can be configured to output more than one comparison data set and / or more than one identifier. This makes it possible to output multiple comparison data sets and / or multiple equivalent identifiers with regard to the application situation.
[0021] The system can have at least one mapping matrix for the 1:n assignment of property values of an application situation. Here, n can preferably be >=2. The mapping matrix can therefore be designed such that one or more property values are assigned to a property value. For example, with regard to a specific type of protective element, for example in the area of fire protection elements, the material of a substrate may be of only limited relevance for selecting a suitable protective element. For example, it could be the case that a fire protection sealant certified for use on concrete is also suitable for use on plasterboard, but not for use on wood. In this case, the mapping matrix could have an assignment of the material property value "plasterboard" to "plasterboard" and "concrete", i.e. a 1:2 assignment.Furthermore, it could simply show the assignment of the material property value “wood” to “wood”, i.e. a 1:1 assignment, etc.
[0022] With the help of such a mapping matrix, suitable comparison data sets can be pre-selected very efficiently for individual properties of the application situations.
[0023] The system can also have multiple mapping matrices, especially for more than one property of the application situations. Sequential filtering based on the various mapping matrices allows for particularly extensive pre-selection.
[0024] The mapping matrices can be obtained with the help of expert assessments.
[0025] The system may comprise a machine learner. In particular, the system may comprise a machine learner and / or be a machine learner. The machine learner may, for example, comprise a neural network, a support vector machine, a deep learning unit, or the like. The machine learner may be trained based on manually prepared expert assessments of application situations and suitable protective elements, for example, the protective elements certified for the respective application situations. It is conceivable that the system, in particular the machine learner, is configured to receive feedback and, particularly preferably, to update, in particular retrain, the machine learner with the aid of the feedback.
[0026] The system can thus be configured to filter one or more of the comparison data sets based on their similarity to the input data set.
[0027] If the similarity is too low, in particular if a limit value of a distance between the input data set representing the application situation and one or more, in particular each, of the comparison data sets is exceeded, the system can be configured not to output a comparison data set, not to output a reference to at least one of the comparison data sets and not to output an identifier.
[0028] In this case, for example, an error signal may be issued. This error signal may indicate that no sufficiently similar comparison data set could be found.
[0029] The type of protective element can include fire protection elements, insulation elements, i.e. elements for protection against energy loss and / or energy input, in particular against heat and / or cold loss, noise protection elements and / or moisture protection elements.
[0030] Particularly in fire protection, special application situations often arise that cannot be covered by standard approval procedures. The system is also particularly advantageous when the protective element type involves fire protection elements, as missing or incorrect fire protection measures can have particularly devastating consequences in the event of a fire.
[0031] The scope of the invention further includes a method for determining a protective element of a protective element type for use in an application situation relating to a building part, wherein, using a system of the type described above and / or below, an input data set comprising data relating to an application situation is compared with a plurality of comparison data sets, each comprising comparison application situations and each having at least one associated protective element of the protective element type, and wherein, as a result of the comparison, at least one of the comparison data sets, a part of at least one of the comparison data sets, a reference to at least one of the comparison data sets and / or an identifier of a protective element of the protective element type is determined.In analogous argumentation as described above for the system, the method also enables particularly cost-effective and safe protection of building components, especially entire buildings.
[0032] Further features and advantages of the invention will become apparent from the following detailed description of exemplary embodiments of the invention, based on the figures of the drawing, which show details essential to the invention, and from the claims. The features shown therein are not necessarily to scale and are presented in such a way that the special features of the invention can be clearly seen. The various features can be implemented individually or in combinations in variants of the invention. Exemplary embodiments of the invention are shown in the schematic drawing and explained in more detail in the following description.
[0033] They show:
[0034] Fig. 1 a building section with different protective elements,
[0035] Fig. 2a-c schematic representations of an application situation, a comparison data set and a result data set;
[0036] Fig. 3 shows a scheme for a method for determining a protective element of a protective element type for a building part,
[0037] Fig. 4 shows a system with several system parts for determining a protective element, Fig. 5 shows a similarity matrix;
[0038] Fig. 6 another similarity matrix and
[0039] Fig. 7 alternative system parts for determining a protective element.
[0040] In the following description of the figures, the same reference numerals are used for identical or functionally corresponding elements to facilitate understanding of the invention.
[0041] The exemplary embodiments described below relate to fire protection measures. Thus, the following statements relate to protective elements designed as fire protection elements. This is to be understood merely as an example; it is conceivable that the systems and methods shown could also be used for other types of protective elements, for example, one or more of the aforementioned types of protective elements.
[0042] Fig. 1 shows a structural component 10 in a schematic, perspective, partially sectioned view, in particular a section of a building, with a plurality of protective elements 12 in different application situations 14. The protective elements 12 are designed as fire protection elements. In particular, they are designed to prevent the spread of fire from one room 16 to another room 16 or vice versa, or at least to impede it, depending on the protection class.The application situations 14 can differ, for example, with regard to the type of substrate material, in the case of wall openings with regard to the dimensions and / or shapes of the wall openings, the elements passing through the wall openings, for example supply lines or power lines, with regard to the required flexibility, i.e. whether or to what extent subsequent changes to the protective element 12 are necessary or possible, with regard to the required fire protection classes to be achieved and / or the like.
[0043] To ensure optimal fire protection, the protective elements 12 for such application situations 14 must be selected and properly installed in accordance with the respective certifications of the protective elements 12. In cases where an application situation 14 does not correspond to any of the application situations for which protective elements 12 are available, in particular certified, a comparable application situation for which a protective element 12 is available should be identified, as far as possible.
[0044] Fig. 2a, 2b and 2c show schematic representations of application situations 14. In the examples shown, these representations include image data 20, nominally scaled data 22 and structured, semi-structured and / or unstructured text data 24. The text data 24 can also be at least nominally scaled.
[0045] The image data represent, among other things, geometric features of the application situations 14.
[0046] The nominally scaled data 22 concern, for example, material designations or protection classes to be achieved. In addition to truly nominally scaled data, they can also include higher-scaled data, particularly ordinally scaled data.
[0047] The text data 24 includes supplementary information about the application situations 14, for example, regarding the layer structure of the floor, ceiling, or wall depicted in the image data 20. It may also contain additional dimensions, references to applicable norms and standards, or the like.
[0048] Fig. 2a shows an example of an application situation 14 from a list of standard application situations.
[0049] Fig. 2b shows an example of an application situation 14 for which a protective element 12 in the form of a fire protection element is certified. In the example shown, this is a tube of fire-retardant sealing paste for sealing gaps between a wall or floor and a pipe passing through the wall or floor. The application situation 14, together with the associated protective element 12, which is certified for the application situation 14, can form a comparison data set 26 for a system described in more detail below. Here and below, application situations 14 of comparison data sets 26 are also referred to as comparison application situations 27.
[0050] Fig. 2c shows an example of a possible result data set 28, which can be obtained, for example, by the system to be described below. In the example shown, the result data set 28 also includes an application situation 14 and an associated protection element 12, which is determined as described in more detail below.
[0051] In the examples according to Fig. 2b and 2c, the comparison data set 26 and the result data set 28 largely correspond, for example, the respective image data 20 match, but differences exist with regard to the text data 24.
[0052] Fig. 3 shows a method 100 for determining a protective element 12 (see Fig. 1) of a protective element type, in the present example in particular a fire protection element, for use in an application situation 14 (see e.g. Fig. 2a) which relates to a building part 10 (see Fig. 1), using a system 30 which in this embodiment comprises a first, a second and a third system part 32, 33 or34, an input data record 36, which comprises data on the application situation 14, is compared with a plurality of comparison data records 26 stored in a database 25, each of which comprises data on comparison application situations 27 and each at least one associated protective element 12 of the protective element type, and wherein as a result of the comparison at least one of the comparison data records 26, at least a part of one of the comparison data records 26, a reference to at least one of the comparison data records 26 and / or an identifier of a protective element 12 of the protective element type is determined.
[0053] Thus, in this embodiment, the system 30 is constructed in several parts. Accordingly, in the embodiment shown in Fig. 3, the method 100 is implemented in several stages.
[0054] In the system 30, in particular in the first system part 32, the input data set 36 is input into an input device 37 of the system 30. The input device 37 can, for example, comprise a data interface with which data can be retrieved online or entered online, for example, on a web-based form. It can also have a human-machine interface, for example in the form of a keyboard, a mouse and / or a touchscreen or the like.
[0055] The first system part 32 pre-filters one or more of the comparison data sets 26 based on the input data set 36. In the example shown in Fig. 3, this results in three different comparison data sets 26.
[0056] The second system part 33 makes a final selection within the pre-filtered comparison data sets 26 and delivers a result data set 38, which in this embodiment thus corresponds to one of the pre-filtered comparison data sets 26.
[0057] The third system component 34 forms a quality assurance level. Based on criteria that are, if possible, independent of the filtering or selection procedures of system components 32, 33, the third system component 34 performs a final plausibility check of the result data set 38. For example, the third system component 34 can be configured to check whether the absolute exclusion criteria specified in the input data set 36 are actually met by the result data set 38. For example, if the input data set 36 concerns a wall opening and a fire protection element is to be determined, the wall thickness of the opened wall can be specified. In this case, the fire protection element corresponding to the result data set 38 should actually be certified for this wall thickness.In the case of a fire protection sealing paste for filling a gap, it can also be checked, for example, whether the fire protection sealing paste is actually certified for the dimensions of a gap to be sealed specified in the input data set 36.
[0058] If this quality assurance is positive, an identifier 42 of the protective element 12 corresponding to the result data set 38 is output on an output device 40, for example. The output device 40 can comprise an image output unit, for example a monitor, a display unit of a computer, for example a tablet or a smartphone, or the like. The output device 40 can alternatively or additionally also comprise a data interface for transmitting the result data set 38 or a portion thereof to another computer. The other computer can, for example, be and / or comprise a planning system for building design. The other computer can also be and / or comprise a construction execution system and / or form part of such a construction execution system.The construction execution system can, for example, be configured to order a sufficient quantity of protective elements 12 of the type corresponding to the identifier 42, to deliver them to a construction site where the building component 10 is being erected, and / or to use them there, in particular to install them, in an automated manner, for example with the aid of a suitable construction robot.
[0059] In general, it is conceivable that the system 30 is also designed in one part. Alternatively, it can also be designed in two parts or more than three parts. Accordingly, the method 100 can alternatively be designed in one step, two steps, or more than three steps. It is also conceivable that the system 30 is designed as a recommendation system and / or the method 100 is designed as a recommendation process. For example, for legal reasons, it may be necessary that a final selection of a protective element cannot be made purely mechanically. In this case, it can be provided, for example, that the results of the system 30 merely constitute a suggestion that is subsequently reviewed by a human expert, particularly in the field of the type of protective element in question. For example, the third system part 34 can be replaced by the human expert for this purpose.
[0060] Fig. 4 shows details of an embodiment of the system parts 32, 33 as well as the corresponding steps of the method 100 (see Fig. 3).
[0061] In this embodiment, the first system part 32 has a plurality of mapping matrices 44 for a 1:n assignment of property values of different properties.
[0062] An example of such a mapping matrix 44 for property values relating to fire protection elements for penetrations is shown in Fig. 5.
[0063] Depending on the marked fields, one or more of the columnar property values A, B, C, D, E, F, G, H, I, J, K, L, M from comparison data sets 26 (see Fig. 4) are assigned to one of the property values A, B, C, D, E, F, G, H, I, J, K, L, or M listed row by row in the input data set 36 (see Fig. 4). For example, if the property relates to a type of substrate material, the property values A, B, C, D, E, F, G, H, I, J, K, L, M can correspond to “concrete,” “plasterboard,” “wood,” “chipboard,” “metal-clad concrete,” etc. For example, the property value C is assigned the property values I and M in addition to the identical property value C. This results in a 1:3 assignment.
[0064] In addition to the identical property value A, property value B is also assigned to property value A. This results in a 1:2 assignment.
[0065] An alternative embodiment of a mapping matrix 44 is shown in Fig. 6. This mapping matrix 44 largely corresponds to the previously described mapping matrix 44. A special feature of this mapping matrix 44, however, is that it also takes into account graduated degrees of assignment, represented in Fig. 6 by different fill levels of the individual fields. The degrees of assignment can, for example, correspond to degrees of similarity. They can also correspond to proportions in which the various property values, in this case material types, should be considered as substitutes.
[0066] As further schematically illustrated in Fig. 4, the first system part 32 is configured to assign further alternative values to the relevant property values for each property that can be isolated from the input data set 36. From the comparison data sets 26, the first system part 32 then filters those comparison data sets 26 that, for each of the isolable properties, each contain either a property value that matches the respective associated property value of the input data set 36 or one of the alternative values.
[0067] To further reduce the number of preselected comparison data sets 26, the first system part 32 can be configured to perform a hierarchical cluster analysis, for example, based on a Gower distance metric, among the retrieved comparison data sets 26. As a result, the first system part 32 can further be configured to select specific clusters and specific elements within them. For example, the first system part can be configured to select the three clusters and their central elements whose central elements are closest to the input data set 36 in terms of the Gower distance metric.
[0068] Thus, for example, three comparison data sets 26 can result as an intermediate result and serve as input data for the second system part 33. In this exemplary embodiment, the second system part 33 has an image processing unit 46. The image processing unit 46 has a machine learner 48 in the form of a multi-stage neural network. The machine learner 48 is trained to segment the image data 20 (see Figs. 2a, 2b, and 2c). For example, it can be configured to distinguish between straight lines and curves, to recognize and / or classify angles, and / or to identify other elements typical for the application situations 14 (Fig. 1), for example, pipelines, supports, building parts, or the like.
[0069] The second system part 33 is configured to classify, with the aid of the image processing unit 46, the image data 20 of the input data set 36 and the image data 20 of the comparison data sets 26 identified as intermediate results. The comparison data set 26 or those comparison data sets 26 that exhibit the greatest similarity according to this classification of their image data 20, for example, again determined using a Gower distance metric, then form the result data set 38 or the result data sets 38.
[0070] Fig. 7 shows a further example of the first and second system parts 32, 33 and the corresponding substeps of the method 100 (see Fig. 3).
[0071] The second system part 33 can correspond to the second system part 33 described in connection with Fig. 4.
[0072] In this embodiment, image data 20 (see Figs. 2a to 2c) of the input data set 36 and of the comparison data sets 26 are again evaluated by the second system part 33. The data extracted from the image data 20 form further, at least nominally scaled, properties of the input data set 36 and the comparison data sets 26.
[0073] The first system part 32 of this embodiment has a further machine learner 48, which, for example, also comprises a multi-layer neural network. The machine learner 48 is trained, for example by means of backpropagation and, for example, using expert assessments, to assign a most similar one of the comparison data sets 26 or the most similar one of the comparison application situations 27 to an input data set 36 and to output the corresponding comparison data set 26 and / or its identifier. The input data set 36 is thus first analyzed with regard to its image data 20. The additional, at least nominally scaled properties obtained in this way are fed to the first system part 32 with the remaining, at least nominally scaled properties of the input data set 36. This, in particular its machine learner 48, determines the most similar one of the comparison data sets 26 orthe associated comparison application situations 27 as a result data set 38 and outputs this on the output device 40.
[0074] As an alternative to the output on the output device 40 or in addition thereto, it is also conceivable that a final quality assurance takes place, for example with the aid of a third system part 34 (see Fig. 3).
[0075] List of reference symbols
[0076] 10 building sections
[0077] 12 protective element
[0078] 14 Application situation
[0079] 16 rooms
[0080] 20 image data
[0081] 22 nominal scaled data
[0082] 24 text data
[0083] 25 Database
[0084] 26 Comparison data set
[0085] 27 Comparison application situation
[0086] 28 Result data set
[0087] 30 systems
[0088] 32 first system part
[0089] 33 second system part
[0090] 34 third system part
[0091] 36 Input data set
[0092] 37 Input device
[0093] 38 Result data set
[0094] 40 Output device
[0095] 42 identifiers
[0096] 44 mapping matrix
[0097] 48 machine learners
[0098] 100 procedures
[0099] A, B, C, D, E, F, G, H, I, J, K, L, M Property value
Claims
Patent claims 1. A system (30) for determining a protective element (12) of a protective element type for use in an application situation (14) relating to a building part (10), comprising an input device (37), an output device (40) and a database (25), wherein the database (25) comprises a plurality of comparison data sets (26), each comprising data relating to a comparison application situation (27), wherein the system (30) is configured to receive, via the input device (37), an input data set (36) comprising data relating to the application situation (14), and, depending on the input data set (36), to output, via the output device (40), at least one of the comparison data sets (26), a part of at least one of the comparison data sets (26), a reference to at least one of the comparison data sets (26) and / or at least one identifier (42) of a protective element (12) of the protective element type.
2. System (30) according to the preceding claim, characterized in that the database (25) comprises at least one identifier (42) of an associated protective element (12) of the protective element type for at least one, preferably for all, of the comparison data records (26).
3. System (30) according to one of the preceding claims, characterized in that the input data set (36) and / or at least one of the comparison data sets (26) comprises image data (20).
4. System (30) according to one of the preceding claims, characterized in that the input data set (36) and / or at least one of the comparison data sets (26) comprises nominally scaled (22) data.
5. System (30) according to one of the preceding claims, characterized in that the input data set (36) and / or at least one of the comparison data sets (26) comprises text data.
6. System (30) according to one of the preceding claims, characterized in that the system (30) is configured to determine a subset of the comparison data sets (26) depending on the input data set (36).
7. System (30) according to one of the preceding claims, characterized in that the system (30) has at least one mapping matrix (44) for the 1:n assignment of property values (510, 511, 520, 530, 540, 541, 550) of an application situation (14).
8. System (30) according to one of the preceding claims, characterized in that the system (30) has a machine learner (48).
9. System (30) according to one of the preceding claims, characterized in that the system (30) is configured to output no comparison data record (26), no reference to at least one of the comparison data records (26) and no identifier (42) when a limit value of a distance between the input data record (36) representing the application situation (14) and one or more of the comparison data records is exceeded.
10. System (30) according to one of the preceding claims, characterized in that the type of protective element relates to fire protection elements, insulation elements, noise protection elements and / or moisture protection elements.
11. Method (100) for determining a protective element (12) of a protective element type for use in an application situation (14) relating to a building part (10), wherein, using a system (30) according to one of the preceding claims, an input data set (36) comprising data relating to the application situation (14) is compared with a plurality of comparison data sets (26), each comprising comparison application situations (27) and at least one associated protective element (12) of the protective element type, and wherein, as a result of the comparison, at least one of the comparison data sets (26), a part of at least one of the comparison data sets (26), a reference to at least one of the comparison data sets (26) and / or an identifier (42) of a protective element (12) of the protective element type is determined.