GB2574582A

GB2574582A - Method for simulating a technical device

Info

Publication number: GB2574582A
Application number: GB1808685.0A
Authority: GB
Inventors: Kirchner Eckhard; Rainer Gabriel Schweiger Martin
Original assignee: Technische Universitaet Darmstadt
Current assignee: Technische Universitaet Darmstadt
Priority date: 2018-05-28
Filing date: 2018-05-28
Publication date: 2019-12-18
Also published as: GB201808685D0

Abstract

A method for generating a sequence of words describing a technical device is provided. In particular the method is used in drafting a patent application. The method comprises providing a computer model of the technical device, and arranging a sequence of words based on a list of components that constitute the technical device, cooperation data of the device and component hierarchy data (see fig 5). The method may comprise retrieving a function of at least one component from a component function database, which can be based on machine learning which based on analysis of device drawings.

Description

Method for simulating a technical device

The present invention relates in general to a method of simulation a technical device, and in particular to a method for simulating a technical device by generating a sequence of words .

With the steadily growing number and complexity of technical devices, the number and the complexity of the technical documentation of the technical devices grows as well. Many people, however, have difficulties to read technical documentations, in particular technical drawings.

The present specification provides an improved method for interpreting computer models of technical devices.

A method for simulating a technical device on a computer is provided. The computer comprises a memory which stores a computer model of the device, a list of components of the device, cooperation data of the device, and component hierarchy data. In a step of the method, a sequence of words is arranged automatically based on hierarchy entries of the component hierarchy data, on components of the list of components of the device, and on cooperation entries of the cooperation data of the device. The sequence of words is output automatically.

Simulating a technical device relates to generating a description of the device and its function which is comprehensible for a human. In typical modern engineering work, a device is simulated repeatedly after a computer model has been created. Besides, for example, fault state simulations or force distribution simulations, the function of the device is predicted according to a textual representation of the device components and their functions. Such a textual representation can be generated manually based on drawings generated from the computer model. The simulation of a technical device according to the present specification constitutes a technical purpose for a computer-implemented method. The method according to the present specification is also functionally directed to that technical purpose. In particular, a sequence of words constitutes a defined class of technical items, the generation of which is a functional technical feature. In the present case, the purpose of simulating a technical device by generating a sequence of words is established in the further steps of the disclosed method. Further, on the basis of the present specification, it is verifiable that the generated sequence of words actually provides a description of a drawing of a technical device, whereby the independent method claims are directed to the simulation of a technical device. The simulation performs technical functions typical of modern engineering work. It provides for realistic prediction of the performance and function of a technical device depicted in the drawing and thereby ideally allows it to be developed so accurately that a prototype's chances of success can be assessed before it is built.

According to the specification, the computer comprises a memory. The memory can comprise any memory suitable to provide data to a processing unit of the computer. Amongst the technologies usable for the memory are, for example RAM, ROM, HDD, SSD, UFS, floppy disk, CDROM, flash memory, or ECC memory.

The memory stores a computer model of the device. The computer model of the device can comprise one or more 2D CAD models, one or more 3D CAD models, one or more circuit diagram, and/or further models of the device. For example, a computer model of a mechanical device with an electrical control can comprise both a 3D CAD model and a circuit diagram.

A 2D CAD model comprises a vector-based representation of the geometry of the device or the geometry of parts of the device in two space dimensions. Therein, a finite number of perspectives of the device can be given. For a simple device geometry, this can be enough to represent the device without ambiguities .

A 3D CAD model comprises a vector-based, surface-based or solid-based representation of the geometry of the device or the geometry of parts of the device in three space dimensions. A vector-based 3D CAD model can comprise a wire mesh. A surface-based 3D CAD model comprises one or more surfaces made up of basic geometries and/or using a free form design. A number of surfaces are combined to define the outer boundaries of the geometry of the device. For example, a surface may be a spherical surface, or a complex organic surface. A solid 3D CAD model comprises a closed 3D representation of the device. The geometry of the device is described using mathematical principles. An object of a solid 3D model is able to be seen in geographic terms, and is considered a watertight model because the internal details of the product can be included. Each part of the model is added one at a time until the complete model is finished. Assembly modeling can also play a part in solid modeling, which is simply when smaller parts of the model make up the whole. Solid models can intersect, join and subtract objects from one another to create the desired results as far as shape and form.

A circuit diagram is also called circuit plan, circuit layout, electronic schematic, wiring diagram, schematic circuit diagram, printed circuit board design, circuit board design, or simply circuit design. A circuit diagram comprises electronic components and the connections between them. Therein, a positioning between the electronic components is provided. The sizes and shapes of the electronic components may be represented perfectly or not. The parameters of the electronic components are provided by the circuit diagram, either directly or in combination with a component catalogue.

The memory of the computer further stores a list of components of the device, cooperation data of the device, and component hierarchy data. The list of components may be a CAD (Computer Aided Design) based list saved in a corresponding CAD data base. The list of components comprises a name of each component, and a corresponding reference numeral.

Cooperation data of the device is for example provided by adding one or more cooperation reference numerals to an entry of a component in the list of components. Cooperation data can further be provided by adding one or more cooperation modes to an entry of a component in the list of components.

Hierarchy data can, for example, be provided by adding one or more superordinate group features to an entry of a component in the list of components. For example, an entry of the list of components may comprise a name 'input shaft', a reference numeral '4', a cooperation reference numeral '5', a cooperation mode 'drives, and a superordinate group feature 'input section' . This would represent, for example, that the input shaft of a device is identified by the reference numeral 4, drives another component which is identified by the reference numeral 5, and belongs to the input section of the device. The device, as a whole, can be also considered as a superordinate group feature, namely as the top-level superordinate feature. The method also comprises a step of retrieving a function of at least one component from a component function database. In particular, one or more component functions of the at least one component may be retrieved from the component function database.

The cooperation data can also be provided in a different form. For example, the cooperation data can be provided in a matrix with a first index for each of the components of the device and a second index for each of the components of the device. Each entry of the matrix specifies whether the two components have a relation with each other and, if so, which relation they have with each other. For example, an entry 34,5 of a matrix (ai,j) is 'drives' . This would represent that the component identified by reference numeral 4 drives another component identified by reference numeral 5. Accordingly, an entry a_5/z of the matrix (a_1;j) should be 'is driven', since the component identified by reference numeral 5 is driven by another component identified by reference numeral 4. The cooperation data and the hierarchy data can be provided in any suitable form or representation similar or different to the examples provided in this specification.

The list of components, the cooperation data and the hierarchy data, in general, comprise the name of a component, the function of the component from the component function database, the at least one cooperation feature of each component, and the at least one superordinate group feature of the component.

Arranging a sequence of words refers to generating a piece of text in any given language. In many languages, arranging a sequence of words will comprise generating one or more sentences or fragments of sentences. Therein, the words which are arranged can be selected from a dictionary of the respective language. When a sequence of words is arranged for a component of the device, words from the respective entry in a components list can be utilized. Alternatively or additionally, the words which are arranged can comprise synonyms, paraphrases or expositions of the words from the respective entry in the components list. Further, the arranged words may comprise descriptions of components, cooperation modes or sets thereof which have a substantially equivalent technical effect to that of the respective components, cooperation modes or sets thereof in the list of components and the cooperation data.

For each component the arranged words may comprise a description of alternative components, of an umbrella or subordinate term for the component and/or of a general principle of the component.

Alternatively or additionally, for each connection of a number of components the arranged words may comprise a description of alternative connections, of an umbrella term for the connection and/or of a general principle of the connection. Therein, a number of components are for example two components, three components, or any other number of components. Examples of a connection of components are cogwheel meshing, a shaft-hub-connection, or a positive-fit connection. A connection can be indicated by cooperation data, by hierarchy data, or by both cooperation data and hierarchy data.

By outputting the sequence of words under taking into account the name of the component, the function of the component from the component function database, the at least one cooperation feature of each component, and the at least one superordinate group feature of the component a description of the technical device can be created which can be understood by a person, even if he has difficulties to interpret the drawing of the technical device. Furthermore, in this way, a detailed description of the technical device can be obtained which can be readily implemented or embedded in a patent, a specification of a patent application or any other technical documentation.

The outputting may comprise outputting the sequence of words, for example, into a template, into a framework, into a file, or into a database. For example, a template may comprise further text which, together with the generated sequence of words, makes for a patent application or a technical documentation. As another example, a framework may distribute the generated sequence of words to an engineer for assessing the technical device which is described with it.

The method may further comprise retrieving a function of at least one component based on a name of the at least one component from a component function database. The cooperation data comprises at least a cooperation feature for each component within the technical device. In particular, the cooperation feature of the component within the technical device can serve as a useful indicator of the function of the component characteristic for the given component.

The retrieval of a function of the at least one component may be based on a likelihood, in particular on a likelihood that the function can be executed by the component, based on earlier learning or previous learning material. The earlier learning may be based on analysis of technical drawings of a plurality of different technical devices. The earlier learning may comprise a machine learning using pattern recognition, deep learning, and/or neural/network learning, based on a large number of earlier technical drawings, in particular based on one or more techniques used in big-data analysis. By using the big data analysis techniques, and by increasing the statistical basis of the data, the likelihood of correct assignment of the function to the component can be increased. The technical significance of this method increases with the speed and capabilities of computer simulations the simulation method, as this enables a more accurate big-data analysis for virtually testing and examining for suitability before the expensive device fabrication process starts.

In one embodiment, the retrieval of the component functions comprises retrieving an alternative component, which has the same or a similar technical effect as the at least one component. The alternative component is retrieved from a component function database. An alternative component can perform the same function as one of the components. For example, a belt drive performs the same function as a chain drive, and has a similar technical effect. In some instances, one alternative component has a similar technical effect as a number of components, and the alternative component could replace a number of components in a technical device. In other instances, a number of alternative components together have a similar technical effect as one component, and the number of alternative components could together replace one component in a technical device.

The list of components may be organized in a hierarchical order, such that each component has at least one immediate superordinate group feature. By taking into account the superordinate group feature of the components, the outputted sequence of words reflects the hierarchical order or technical context of the components, facilitating the correct interpretation of the sequence of words as the description of the device .

In the sequence of words, the words may be arranged and syntactically connected in such a way that the sequence of words reproduces at least one technical subject matter related to at least one superordinate group feature. Thus, the sequence of words can provide a partial description related to any superordinate group feature or section of the device. This can be helpful for a detailed analysis of any particular section of the device.

The sequence of words may be arranged in the form of the names of the components, followed by corresponding reference numerals, according to the names of the components, wherein the order of the words takes the function of the component retrieved from the function database, the at least one cooperation feature of each component, and the at least one superordinate group feature of the component into account. Such an arrangement of the sequence of words is particularly helpful for providing a so-called picture description or structural description of the technical drawing.

The order of the words can further take an alternative component into account, which has the same or a similar technical effect as at least one of the components. An alternative component may have the same effect as the at least one component when it performs the same function as the at least one component. The alternative component can be retrieved from a component function database. The alternative component can also be provided with the computer model of the technical device. Alternatively or additionally, the alternative component is provided, selected, or confirmed by a user, for example in a dialogue accompanying or preceding the simulation.

It is possible to arrange the sequence of words is arranged to form an apparatus claim encompassing at least one of the superordinate group features, apparatus patent claims for a patent application can be easily drafted.

It is possible to arrange the sequence of words in the form of process steps according to the functions of the components, retrieved from the function database, wherein the order of the words takes the at least a part of the cooperation data for each component, and at least a part of the hierarchy data for the component into account. Such an arrangement of the sequence of words can be used for describing an operation of the device or of any superordinate group feature or section of the device, in particular, when the technical subject matter comprises at least one operation method of at least one superordinate group feature of the technical device.

The at least one operation method may comprise a sequence of operation methods, wherein the sequence of operation methods is arranged in a hierarchical order, in accordance to the respective superordinate group features.

The sequence of words may further be arranged in the form of a method claim encompassing at least one operational method of at least one superordinate group feature. In this way, method claims for a patent application can be easily drafted.

The sequence of words arranged in the form of a patent claim may comprise a preamble or generic part and a characterizing portion or portion specific for the subject matter with regard to the previous learning material.

According to another aspect, a method for drafting a patent application is provided which is at least partially based on the method for simulating a technical device by generating a sequence of words describing a drawing of a technical device according to the first aspect. In particular, the method for simulating a technical device by generating a sequence of words describing a drawing of a technical device can be used for drafting a description and/or the claims of the patent application.

The drafting of the patent application can be, thus, at least partially carried out by a computer or a patent drafting robot, resulting in a increased throughput of patent drafting.

In the following description, details are provided to describe the embodiments of the application. It shall be apparent to one skilled in the art, however, that the embodiments may be practiced without such details.

Fig. 1 show a structure of a patent application,

Fig. 2 shows an order of drafting a patent application according to an embodiment,

Fig. 3 shows a gearbox as an example of a technical device,

Fig. 4 shows the gearbox of Fig. 3 with indication of sections functional sections,

Fig. 5 shows a list of components with reference numerals in accordance to Fig. 4,

Fig. 6 shows a circuit as an example of a technical device,

Fig. 7 shows a list of components with reference numerals in accordance to Fig. 6, and

Fig. 8 shows a method of generating a sequence of words according to an embodiment.

Fig. 9 shows a graph depicting the development of the information content of a device description over time according to an embodiment.

Fig. 1 shows a typical structure of a patent application. The structure shown in Fig. 1 is an example of a typical structure generally used for patent applications for protecting an invention. In this example, the patent application comprises a title, an introduction or a starting point, a prior art description, a problem which is solved by the invention to be protected, a support section describing a solution of the problem, a figure description, a reference numerals' list, patent claims, figure(s) and abstract.

Depending on patent legislation and patent praxis, there may be deviations from this structure. For instance, according to the patent practice in the United States, the problem section is usually omitted in patent applications. However, all patent applications need to have a description for explaining the invention and patent claims for defining the scope of protection.

A support section usually provides a support to the patent claims and describes a solution of the problem in relatively general or abstract terms. The support section may also provide a link between the figure description, the wording of which is usually more specific in order to explain the invention in an enabling way, and the broader and more generic wording of the patent claim. Often there is a reference numerals list which shows the reference numerals which have been used in the figure description.

Instead of drafting a patent application with the structure shown in Figure 1, starting from the title or from the introduction and then drafting the patent application from the beginning till the end, in the sequence of the structure of Fig. 1, we suggest a specific order of drafting a patent application.

Instead of starting drafting with the claims, we provide a comprehensive picture of the specific subject matter of the embodiments before starting the drafting of the claim or the claims .

Figs. 2 shows an order of drafting a patent application according to one embodiment. As shown in Fig. 2, in this embodiment, the drafting starts by providing a figure or a set of figures, in step 100. The Figures usually represent key points of a patent application, because the figures usually originate from inventors who are using the figures to explain their invention and hence, the figures contain important information about how the invention works. Moreover, in some cases the figures are indispensable to provide an enabling disclosure which is a prerequisite of patentability.

After providing the figures, which can be, in particular, figures produced using a CAD (Computer Aided Design) program, in step 150, the figure description is drafted. In the course of drafting the figure description, a reference numerals list is provided as well. The reference numerals list may be, in particular, also provided by the CAD program which is used for drawing the figures, such that the same numerals can be used also for the figure description.

Starting from the figure description, a set of patent claims is drafted. Hereby, in step 200, a first a set of so-called picture patent claims or picture claims is drafted. The picture claims are based on the figure description and in general contain specific structural features shown in the figures. Usually, picture claims are rather specific and can have a relatively narrow scope of protection.

After drafting a set of picture claims, a set of broad claims and a set of means-plus-function claims can be drafted as well, in step 250. This gives a broad patent protection for the patent if it is granted, since the patent claims define what is protected.

After completing the claims, the remaining sections in accordance with the structure of Fig. 1, in particular, abstract, introduction, support and problem, can be drafted. According to the embodiment of Fig. 2, in step 300, the abstract section is drafted, which is essentially a patent claim put into proper sentences. In step 350, after the abstract is drafted, the introduction of the patent application is drafted.

Thereafter, in step 400, the support section is drafted. The Support Section describes the solution of the problem in the wording of the patent claims. The Support Section is also used to connect the figure description, which can be very specific, with a broad wording of the broad patent claims. In the support section, the effects of the elements of the patent claims is also added for better understanding the invention.

In step 450, after drafting a support section, the prior art description is drafted. For drafting the prior art description, the abstract sections of the prior art documents that are available can be reworded and put into the patent application.

After completing the prior art description, the problem section, which describes problem that is solved by the invention, is drafted, if required. This completes the patent application as shown in Fig. 2.

Figs. 3-8 show the sequence of outputting the figure description and picture claims according to an embodiment of the present application.

Fig. 3 shows a gearbox as an example of a technical device. As shown in Fig. 3, the gearbox has a first shaft 4 and a second shaft 7. There is a first gear wheel 5 on the first shaft 4 and a second gear wheel 6 on the second shaft 7. The gear wheels are meshing in a middle section. An input torque M±_n, which is shown by a first wide arrow directed inwards with respect to the gearbox, generates an output torque M_out shown by a wide arrow directed outwards with respect to the gearbox. In this example, the output torgue M_out would be smaller than the input torque M_in because of the different diameters of the gear wheels. Assuming that the above described gearbox is an embodiment of the invention, a description of a patent application can be provided based on a description of this gearbox .

Fig. 4 shows the gearbox of Fig. 3 with indication of sections functional sections. As shown in Fig. 4, further reference numerals are added to the drawing of the gearbox 1, Further, arrows are added to denominate larger sections. In particular, the gear box '1' has an arrow directed down towards the gearbox, while the reference number for specific parts are simple lines that go to the specific part from the reference numeral for that part. The input section '2' is also a section so it gets an arrow, and the output section '3' also gets an arrow, while the input shaft '4' and the first gear wheel '5' are marked with a simple line, respectively.

Fig. 5 shows a list of components with reference numerals in accordance to Fig. 4. As shown in the table of Fig. 5, the table contains the following columns: reference numerals, section name, part name, cooperates with, and cooperates how. Thus, in addition to the reference numerals and part names and section names, the table also includes additional two columns, because they are useful if a robot is used for drafting the figure description. In particular, based on such a table, the robot would know how to describe the cooperation between the various parts and sections. For example, it can be seen from Fig. 4 that the input section 2 cooperates with a reference numeral 3 which is the output section. Further, the entry in the column cooperates how shows that the input section 2 drives the output section 3.

The robot will thus know that there is a gear box 1, the gear box having an input section which drives the part number '3' which is the output section and the output section cooperates with a part number '2' which is the input section, and it is driven by the input section. The situation is the same for the output section: the output section '3' has a second gear wheel '6' which is connected to the output shaft '7', and the output shaft '7' cooperates with the part number '6' which is the second gear wheel and it is driven. As a result, the input torque M_in cooperates with an input section and it is inputted in the direction of the 'Mi_n' arrow shown, and the output torque M_out cooperates with the part number '3' which is the output section and it is outputted in the direction of the 'Mout' arrow shown. Once everything comes together, the figure description can be drafted.

Thus, the table shown in Fig. 5 can be used by the robot to draft a figure description, essentially following the structure of the table.

The figure description is starting with the sentence: the figure shows a gear box '1' with an input section '2' that drives the output section '3'. Thus, the figure description starts on a section level or on an upper level of the gearbox structure. For a more complex figure, with more levels, this would be the highest section or top level. For the gearbox of Fig. 4, the structure of the table of Fig. 5, the next sentence of the description should describe that the output section '3' is driven by the input section '2' . In some embodiments, the robot may skip such sentences if the robot recognizes clear redundancies.

After finishing the first section level description, the robot goes one level deeper and describes the composition of each section, e.g. in the following words the input section '1' comprises an input shaft '4' that is connected with a first gear wheel '5'. Remaining on the same level, the next few sentences will then be: the output section '3' comprises a second gear wheel '6' that is connected to an output shaft '7'. The first gear wheel '5' meshes with the second gear wheel '6'. An input torque M_in that is input into the input section '2' is converted into an output torque M_out from the output section '3'. This completes the picture description of the gearbox example, corresponding to step 150 of Fig. 2.

The robot further adds sentences describing embodiments of the technical device which are equivalent in function to the embodiment presented in the computer model. The robot adds a sentence as follows: The input shaft '4' is connected with the first gear wheel '5' via a shaft-hub connection. In one embodiment, the input shaft '4' is connected with the first gear wheel '5' via a form-fit connection. In another embodiment, the input shaft '4' is frictionally engaged with the first gear wheel '5'. In another embodiment, the input shaft '4' is firmly bonded with the first gear wheel '5'. In another embodiment, the input shaft '4' and the first gear wheel '5' are forged in one piece.

In one embodiment which is not displayed in the figures, the robot asks a user questions for identifying suitable alternatives or filtering out unsuitable alternatives. For example, the robot asks a user whether a bevel gear should fall under the scope of the description. If the user answers that a bevel gear should fall under the scope of the description, the robot adds the sentence The first gear wheel '5' meshes with the second gear wheel '6'. The input shaft '4' and the output shaft '7' may be angled towards each other. In this case, the first gear wheel '5' and/or the second gear wheel '6' may be helically toothed. Alternatively, the first gear wheel '5' and the second gear wheel '6' may be connected by a friction drive. If the user answers that a bevel gear should not fall under the scope of the description, the robot adds the sentence The input shaft '4' is parallel to the output shaft '7'. The first gear wheel '5' meshes with the second gear wheel '6'. Alternatively, the first gear wheel '5' and the second gear wheel '6' are connected by a chain drive. Alternatively, the first gear wheel '5' and the second gear wheel '6' are connected by a belt drive. Alternatively, the first gear wheel '5' and the second gear wheel '6' are connected by a friction drive..

The robot accesses a database of components with their respective alternatives, cooperation modes with their respective alternatives, and connections with their respective alternatives. Alternatively or additionally, a user may add alternatives for components, cooperation data, or connections to the computer model data. The robot adds sequences of words describing each alternative which is labeled suitable according to a user's answers to a number of questions. For example, each alternative can be stored in a database together with a Boolean term, wherein the answer to each of the number of questions is represented by a Boolean variable. If the Boolean term stored together with an alternative equals true after the number of questions are answered by the user, the alternative is labeled suitable and described by the robot.

Therefore, alternative embodiments having the same or a similar technical effect as the original embodiment represented in a computer model can be described. The sequence of words is then better suited for simulating the technical device in all its breadth.

Referring again back to Fig. 2, after completing step 150, a picture claim can be derived from the figure description. All that is needed is to take the essential elements from the figure description and put them together in the proper claim language. The picture claim would be worded like this: a gearbox 1 with an input section 2 that drives an output section 3, wherein the output section 3 is driven by the input section 2. This sentence is redundant but as it is outputted by a robot, it can be accepted that it is there. This sentence can be struck out once the output text is available. The next sentence of the picture claim would read the input section 1 comprising an input shaft 4 that is connected with a first gear wheel 5, and the next sentence would be one the output section 2 comprising a second gear wheel 6, that is connected with an output shaft 7. The next sentence would be the first gear wheel 5 meshing with the second gear wheel 6, and wherein an input torque M_in that is input into the input section 2, is converted into an output torque M_out from the output section 3. The last sentence provides a functional feature. The picture claim is now complete.

Fig. 6 shows a circuit 10 as an example of a technical device. As shown in Fig. 6, the circuit has an input contact 11 and a grounding 12. An input voltage U_e is applied between the input contact 11 and the grounding 13, which drives the circuit. The circuit further comprises an amplifying section 14, which comprises an operational amplifier 15, which has a noninverting input contact 16 and an inverting input contact 17, which can be called input terminal together. The operational amplifier 15 is driven by a terminal voltage U_PN applied between the non-inverting input contact 16 and the inverting input contact 17. The operational amplifier outputs an output voltage U_a between an output contact 18 and the grounding 12. The output contact 18 is further electrically connected to a feedback section 20. The feedback section 20 comprises a first resistor 21 which separates the output contact 18 and the inverting input contact 17. The circuit also has a second resistor 22 which separates the inverting input contact 17 and the grounding 12. Assuming that the above described circuit is an embodiment of the invention, a description of a patent application can be provided by the patent drafting robot or by the computer based on a description of this circuit.

Fig. 7 shows a list of components with reference numerals in accordance to Fig. 6. As shown in the table of Fig. 7, the table contains the following columns: reference numerals, section name, part name, cooperates with, and cooperates how similarly to the table of Fig. 5 and for the same reasons as outlined in the description of Fig. 5. For example, it can be seen from Fig. 7 that the input section 11 cooperates with a reference numeral 14 which is the amplifying section. Further, the entry in the column cooperates how shows that the input section 11 drives the amplifying section 14. The robot will thus know that there is a circuit 10, the circuit having an input section which drives the part number '14' which is the amplifying section, and the amplifying section cooperates with a part number '20' which is the feedback section, wherein the amplifying section drives the feedback section. Similarly, the robot or the computer will know that the circuit has a feedback section cooperating with a part number 'U_a' which is the output voltage and a part number '15' which is the operational amplifier, wherein the feedback section feeds back the output voltage U_a to the operational amplifier 15. In more detail, the robot will know that the input voltage 'U_e' cooperates with a part number '12' which is an input contact and a part number '13' which is a grounding, wherein the input voltage is applied between the input contact and the grounding. In even more detail, the robot will read from the table of Fig. 7 that the input contact is on the electrical potential of an element 'U_e' which is the input voltage, and that the grounding is electrically connected to a part number '22' which is a second resistor. Further, the robot will know that the second resistor 22 cooperates with the part number '13' which is said grounding 13, and with the part number '17' which is the inverting input contact 17. The robot will know that the second resistor 22 connects the grounding 13 with the inverting input contact 17, while a first resistor 21 connects a part number '18', which is the output contact 18, with the inverting input contact 17. The table further provides the robot with the information that the inverting input contact 17 is driven by the feedback section 20 while the output contact 18 drives the feedback section 20. The robot further will know that the non-inverting input contact is on the electrical potential of the part number 'U_e' which is the input voltage U_e. Accordingly, the robot will know that the element number 'U_PN' which is the terminal voltage is applied between the element number '16' which is the non-inverting input contact 16 and the element number '17' which is the inverting input contact 17. Similarly, the robot will know that the output voltage U_acooperates with the part number '15' which is the operational amplifier 15 and that the output voltage U_a is output by the operational amplifier 15. Once everything comes together, the figure description can be drafted by the patent robot or the computer.

Thus, the table shown in Fig. 7 can be used by the robot to draft a figure description, essentially following the structure of the table.

The figure description is starting with the sentence: the figure shows a circuit '11' with an input section '11' that drives the amplifying section '14'. Thus, the figure description starts on a section level or on an upper level of the circuit structure. For a more complex figure, with more levels, this would be the highest section or top level. For the circuit of Fig. 6, the structure of the table of Fig. 7, the next sentence of the description should describe that the amplifying section '14' drives the feedback section '20' . A third sentence of the description should describe that the feedback section '20' feeds back the output voltage 'U_a' to the amplifying section '14'. In some embodiments, the robot may skip such sentences if the robot recognizes clear redundancies .

After finishing the first section level description, the robot goes one level deeper and describes the composition of each section, e.g. in the following words The input section '10' comprises an input contact '12' that is on the electrical potential of the input voltage 'U_e' . The input section '10' comprises a grounding '13'. Further, the input section '10' comprises an input voltage 'U_e' which is applied between the input contact '12' and the grounding '13' Remaining on the same level, the next few sentences will then be: The amplifying section '14' comprises an operational amplifier '15', which amplifies the input voltage 'U_e' . The amplifying section '14' comprises a non-inverting input contact '16' that is on the electrical potential of the input voltage 'U_e' . The amplifying section '14' further comprises an inverting input contact '17' which is driven by the feedback section '20' and an output contact '18' which drives the feedback section '20'. The operational amplifier '15' comprises a terminal voltage 'U_PN' which is applied between the non-inverting input contact '16' and the inverting input contact '17' . The amplifying section '14' comprises an output voltage 'U_a' which is output by the operational amplifier '15'Once again remaining on the same level, the next few sentences will then be, for instance: The feedback section '20' comprises a first resistor '21', which connects the inverting input contact '17' and the output contact '18'. The feedback section '20' further comprises a second resistor '22' , which connects the inverting input contact '17' and the grounding '13. This completes the picture description of the circuit example, corresponding to step 150 of Fig. 2.

The robot further adds sentences describing embodiments of the technical device which are equivalent in function to the embodiment presented in the computer model. The robot adds a sentence as follows: In one alternative embodiment, the first resistor '21' can be replaced by a load. In another alternative embodiment, the resistor '21' can be replaced by a lightbulb. In another embodiment, the resistor '21' can be replaced by an immersion heater.

In one embodiment which is not displayed in the figures, the robot asks a user questions for identifying suitable alternatives or filtering out unsuitable alternatives. For example, the robot asks a user whether an inverting amplifier should fall under the scope of the description. If the user answers that an inverting amplifier should fall under the scope of the description, the robot adds the sentence The amplifier can alternatively be an inverting amplifier. In this case, the first resistor '21' is placed between the input contact '12' and the inverting input contact '17', the second resistor '22' is placed between the inverting input contact '17' and the output contact '18' , and the non-inverting input contact '16' is connected to the grounding '13'If the user answers that an inverting amplifier should not fall under the scope of the description, the robot adds the sentence The output voltage 'U_a' is in phase with the input voltage 'U_e'

Therefore, alternative embodiments having the same or a similar technical effect as the original embodiment represented in a computer model can be described. The sequence of words is then better suited for simulating the technical device in all its breadth. By asking a user questions about possible alternative embodiments, further information about the technical device is gathered from the user which the user might not have thought about in the first place. Therefore, the information content of the sequence of words can be improved even further.

Referring again back to Fig. 2, after completing step 150, a picture claim can be derived from the figure description. All that is needed is to take the essential elements from the figure description and put them together in the proper claim language. The picture claim would be worded like this: a circuit (10) with an input section (11) that drives an amplifying section (14), wherein the amplifying section (14) drives a feedback section (20)The next part of the picture claim would read the input section (11) comprising an input input contact (12) which is on an electrical potential Ue, a grounding (13), and an input voltage U_e which is applied between the input contact (12) and the grounding (13), and the next part would be the amplifying section comprising an operational amplifier (15) which amplifies the input voltage U_e, a non-inverting input contact (16) which is on the electrical potential of the input voltage U_e, an inverting input contact which is driven by the feedback section (20), an output contact (18) which drives the feedback section (20), a terminal voltage U_PN which is applied between the non-inverting input contact (16) and the inverting input contact (17), and an output voltage U_a which is output by the operational amplifier (15)This part provides the functional feature by describing that the operational amplifier amplifies the input voltage U_e and outputs the output voltage U_a. The next part would be the feedback section (20) comprising a first resistor (21) which connects the inverting input contact (17) and the output contact (18), and a second resistor (22) which connects the inverting input contact (17) and the grounding (13)This part is redundant but as it is outputted by a robot, it can be accepted that it is there. This sentence can be struck out once the output text is available. The picture claim is now complete.

Thus, based on a figure description, a picture claim can be readily drafted. In general, the better structured the figured description is, the easier is to draft the claims. In particular, a well-structured figure description comprises three sections. The first one, is the description of the structure shown in the figures. The second section is a description of assembling the structure shown in the figures. One reason to provide the description of assembly is that, in such way, one would know whether a proper reference numeral has been assigned to each essential part in the figure description, because this is something that one would need. The third section of a good figure description would be a description of operation of the structures shown in the figures. This section of the figure description would provide a basis for method claims. The important rule for drafting a good figure description is as follows: be specific as possible and leave away elements that are not essential for the invention. By following this rule, a very good figure description, which is not too long, can be obtained.

In summary, a preferred order of drafting a patent application starts with the figures, then the figure description and the reference numerals list are done. The important thing is that the picture patent claims can be produced and by doing so, up to 80% of the total work for drafting a patent application will be covered. If the above is done using a patent drafting robot, five times more patent applications can be drafted in the same time as before, when all the above drafting work were done manually. All what is needed are good figures, together with a reference numerals list, and these components are provided by Computer Aided Design (CAD) systems. After the figure description and picture claims are produced, a patent attorney can come in to add broad patent claims and meansplus-function claims, the introduction/ starting point, and also the abstract. The patent attorney can then add the support section or the abstract solution of the problem. The prior art description section is also quite mechanical, which means it can also be done by a robot. The patent attorney then only needs to add a problem which is solved by the invention. Very often this problem is so easy that it can also be done mechanically by a robot.

Fig. 8 shows a method of generating a sequence of words according to an embodiment. According to the method 600 of Fig. 8, in the first step 650, a list of components constituting the is provided. The list of components of components comprises a name of each component, a corresponding reference numeral, at least one cooperation feature of each component with respect to at least one other component of the device, and at least one superordinate group feature for each component. Superordinate group feature denominates a superordinate component or section name of a respective component. In particular, the top superordinate group feature is the device itself.

In step 700, one or more component functions are retrieved from a component function database. The retrieval of the component function is based on the name of the at least one component and the at least one cooperation feature of the component.

In step 750, a sequence of words in arranged in an order which takes the function of the component, the at least one cooperation feature of each component, and the at least one superordinate group feature of the component into account. Thereafter, in step 800, a sequence of words is outputted.

In an embodiment, the retrieval of the component functions is at least partially based on a likelihood that the retrieved function fits to the component to which the function is assigned, the likelihood being based on the results of earlier learning. The earlier learning may comprise, in particular, a machine learning based on analysis of technical drawings of a plurality of different technical devices. The earlier learning may comprise a machine learning using pattern recognition, deep learning, and/or neural/network learning, in particular based on a large number of earlier technical drawings and using big-data analysis technique. By using the big data techniques, with the growing basis of the data, the likelihood of correct assignment functions to the components can be increased.

The list of components can be ordered according to a hierarchical order such that such that each component has at least one immediate superordinate group feature. In this way, it can be ensured that in the sequence of words outputted or generated based on such a list of components, the components are not isolated from their technical context. Furthermore, by taking into account the superordinate group feature of the components, the outputted sequence of words would reflect the hierarchical order of the list, making the sequence of words a realistic description of the device.

According to an embodiment, in the outputted sequence of words, the words are arranged and syntactically connected in such a way that the sequence of words reproduces at least one technical subject matter related to at least one superordinate group feature. In particular, the sequence of words can provide a description of at any superordinate group feature or section of the device, including the device itself as the toplevel superordinate group feature .

In an embodiment, the sequence of words is arranged in the form of the names of the components, followed by corresponding reference numerals, wherein the order of the words takes the function of the component retrieved from the function database, the at least one cooperation feature of each component, and the at least one superordinate group feature of the component into account. In an embodiment, the sequence of words forms an apparatus patent claim encompassing at least one of the superordinate group features of the device.

In an embodiment, the sequence of words is arranged in the form of process steps corresponding to the functions of the components, retrieved from the function database, wherein the order of the words takes the at least one cooperation feature of each component, and the at least one superordinate group feature of the component into account. In an embodiment, the sequence of words forms a method claim encompassing at least one operational method of at least one superordinate group feature .

Fig. 9 shows a graph depicting the development of the information content of a device description over time according to an embodiment. On a vertical axis 901, the information content of a device description such as a technical documentation or a patent application is plotted over the time on a horizontal axis 902. A dashed line represents the information content over time according to a traditional path of a device description, while a dot-dashed line represents the information content over time according to a method of an embodiment of present application.

Therein, at a starting point 910 in time, the technical device is conceived by an inventor as an idea. At this stage, the inventor can have a more or less precise idea of his invention. However, the description thereof is only rudimentary. Subsequently, the inventor will generate a CAD model of the technical device, increasing the information content of the description of the technical device significantly up to the point 911. According to an embodiment of the present invention, the inventor will additionally specify the cooperation of the parts in the CAD model, thereby increasing the information content of the description of the technical device even further up to the point 912.

Now referring back to a traditional development of the description of the technical device, the inventor will generate print-outs, for example print-outs on paper, in image files or in text files, from the CAD model. The inventor can include print-outs from different perspectives and comprising mesh views and/or rendered views in order to convey the technical device in a comprehensible way. The CAD model itself is not included in the description of the technical device at this point, because it is not accepted by the addressee of the description of the technical device, for example a patent office. Therefore, much information is lost in the process, leading to a significantly lower information content of the description of the technical device at the point 913. This is usually considered acceptable because the receiver of the description of the technical device at this point, for example an author of a technical documentation or a drafter of a patent application, can make up for the loss of information in big parts by interpreting the print-outs in the light of his technical expertise. Therefore, in the subsequent process of generation a text description from the print-outs, the information content of the description of the technical device will increase again, up to a point 915. However, the information content of the description of the technical device will necessarily stay lower that what the inventor initially generated when he put everything he has conceived into the CAD model at point 911. When the inventor has bad luck and ends up with a bad patent attorney, for example with a patent attorney who is not fluent in the field of the inventor, the information content can even decrease after point 913, such that at a point 916 almost nothing is left of the original idea of the inventor. Example reasons for a decreasing information content are misunderstandings, oversimplifications, or impracticality due to technical mistakes .

Now referring to the development of the information content of the description of the technical device using a method according to an embodiment of present application, the information content can be mainly preserved. Because a textual description of the technical device is generated directly from the CAD model comprising the cooperation data, only a minor fraction of the information content of the description of the technical device is lost, if any. Every aspect of the CAD model can be meticulously described. Contrary to a human author, a robot would neither forget to describe any elements nor take shortcuts in the description which lead to a reduction in its information content.

In an embodiment, the robot can match the CAD model with a database of reference models. The robot can then identify not only parts which are specified in the CAD model but also parts which are missing in the CAD model as compared to a matching reference model. For example, when a CAD model represents a bicycle with an unusual type of spokeless wheels, the robot can match the CAD model against a reference model of a bicycle. It can then identify that the CAD model does not comprise spokes in the wheels of the bicycle and point out this difference in the textual description. Thereby, the textual description of the technical device is even more precise and complete.

Eventually, the information content of the description of the technical device is not altered further from point 920, where the robot has fully described the computer model, until point 921. This is because the application of the robot is the only step involved in generating of the textual description of the technical device from the CAD model. Therefore, a higher information content is preserved compared to the established method of generating a sequence of words for simulating a technical device.

When alternatives with equivalent or similar function for one or more components are described by the robot, the information content of the sequence of words can rise further after point 920. In particular, when the robot asks a user questions regarding suitable and unsuitable alternatives for components, cooperation modes and connections within the technical device, the information content can, at a point 925, even exceed the information content originally provided by the user at point 912 .

While at least one exemplary embodiment has been presented in the foregoing detailed description, it should be appreciated that a vast number of variations exists. It should also be appreciated that the exemplary embodiment or exemplary embodiments are only examples, and are not intended to limit the scope, applicability, or configuration of the disclosure in any way. Rather, the foregoing detailed description will provide those skilled in the art with a convenient road map for implementing the exemplary embodiment or exemplary embodiments .

Reference symbols and numerals gearbox input section output section input shaft first gear wheel second gear sheel output shaft

Patent application title introduction prior art description problem which is solved by the invention support section reference numerals list patent claims figures abstract drafting a patent application

100 providing figure (s)

150 drafting a figure description with a reference numerals list

200 drafting picture patent claims

250 drafting broad claims and means plus function claims

300 drafting an abstract

350 drafting an introduction

400 drafting a support section

450 drafting a prior art description

500 describing a problem to be solved generating a sequence of words providing a list of components retrieving a function of the components arranging a sequence of words outputting a sequence of words

Mi input torque tout output torque

Claims

1. A method for simulating a technical device on a computer, comprising a memory which stores a computer model of the device, a list of components of the device, cooperation data of the device, and component hierarchy data, comprising the following steps:

arranging a sequence of words based on the hierarchy entries of the component hierarchy data, the list of components of the device, and the cooperation data of the device; and outputting the sequence of words.

2. The method of claim 1, wherein the method further comprises retrieving a function of at least one component based on a name of the at least one component from a component function database.

3. The method of claim 2, wherein the retrieving a function of the at least one component is based on a likelihood based on earlier learning.

4. The method of claim 3, wherein the earlier learning comprises a machine learning based on analysis of a plurality of technical drawings of different technical devices .

5. The method of claim 2 or claim 3, wherein the retrieving comprises retrieving an alternative component, which has the same or a similar technical effect as the at least one component.

6. The method of one of the previous claims 1 to 5, wherein the list of components is arranged in a hierarchical order, such that each component has at least one immediate superordinate group feature.

7. The method of claim 6, wherein in the sequence of words, the words are arranged and syntactically connected in such a way that the sequence of words reproduces at least one technical subject matter related to at least one superordinate group feature.

8. The method of claim 7, wherein the at least one technical subject matter comprises a sequence of technical subject matters arranged in the hierarchical order of the superordinate group features of the drawing.

9. The method of one of the previous claims, wherein the sequence of words is arranged in the form of the names of the components, followed by corresponding reference numerals, according to the names of the components, and wherein the order of the words takes the function of the component retrieved from the function database, the at least one cooperation feature of each component, and the at least one superordinate group feature of the component into account.

10. The method of claim 9, wherein the order of the words further takes an alternative component into account, which has the same or a similar technical effect as at least one of the components.

11. The method of claim 9 or claim 10, wherein the sequence of words is arranged to form an apparatus claim encompassing at least one of the superordinate group features .

12. The method of one of the claims 2 to 8, wherein the sequence of words is arranged in the form of process steps according to the functions of the components, retrieved from the function database, and wherein the

5 order of the words takes at least a part of the cooperation data for each component, and at least a part of the hierarchy data of the component into account.

13. The method of claim 12, wherein the sequence of words is

10 a method claim encompassing at least one operational method of at least a part of the hierarchy data.

14. A method for drafting a patent application, the method being at least partially based on the method for

15 simulating a technical device by generating a sequence of words describing a drawing of a technical device according to one of the claims 1 to 13.