WO2021129922A1

WO2021129922A1 - Method for operating an automation system, and automation system

Info

Publication number: WO2021129922A1
Application number: PCT/EP2019/086909
Authority: WO
Inventors: Andrés Botero Halblaub; Jan Götz; Tim Schenk; Jan Christoph Wehrstedt; Markus Michael Geipel
Original assignee: Siemens Aktiengesellschaft
Priority date: 2019-12-23
Filing date: 2019-12-23
Publication date: 2021-07-01

Abstract

The invention relates to a method for operating an automation system (100), which comprises at least one production plant (110 - 112) for the automatic production of a product (P) with predefined product properties (SPEC) from a number of N reactants (E1 - E3), where N > 1. Formal descriptions are generated in a formal language based on digital twins (D110 - D112) of the production plants (110 - 112), digital twins (DE1 - DE3) of the reactants (E1 - E3) and a digital twin (DP) of the product (P). A manufacturing instruction (PLAN) is determined based on the formal descriptions. The advantage of the method is that flexible and automated production of different products is possible.

Description

description

Method for operating an automation system and automation system

The present invention relates to a method for operating an automation system and an automation system.

Automation systems are known which are suitable for the automated manufacture of products, such as robot factories for example. Such factories can be set up to manufacture different products. A production plan must often be specified explicitly for each individual product. The production plan includes all the steps necessary to manufacture the product, starting with raw materials or semi-finished products through certain intermediate stages or intermediate products to the finished product.

For example, the factory's production facilities have certain capabilities. Based on the capabilities, production steps can be assigned to production systems. Such automation systems do not have a high degree of flexibility, since, for example, products for the production of which new skills are necessary cannot be produced automatically, although one of the production systems would possibly be able to carry out the required skill. Rather, an operator of the automation system would have to implement the new capability manually, which is time-consuming and costly.

An individual production of small numbers of individual products, which differ in each case, as provided in the case of make-to-order, is therefore not efficiently possible. Furthermore, flexible adaptation of the target orientation in production cannot be implemented, since the optimization of production with regard to individual parameters can depend on a large number of dynamic variables. Against this background, it is an object of the present invention to propose an improved method for operating an automation system.

Accordingly, a method for operating an automation system is proposed which comprises at least one production plant for the automated production of a product with predetermined product properties from a number of N starting materials, with N> 1. The product and each starting material can be characterized by a number of property parameters. In a first step a) a digital twin is provided for each production plant of the automation system, a digital twin for each of the N starting materials and a digital twin for the product to be manufactured. In a second step b), a formal skill description is generated for each production plant on the basis of the respective digital twin of the production plant, with the formal skill description showing a formal relationship between a control variable of the production plant and a manipulated variable associated with the control variable a formal language. In a third step c), a formal educt description is generated for each of the N educts on the basis of the digital twin of the respective educt, the respective formal educt description a formal connection between the manipulated variable and the property parameters of the educt in the formal language includes. In a fourth step d), a formal product description of the product is generated on the basis of the digital twin of the product, the formal product description comprising a formal relationship between the predetermined product properties and the property parameters of the product in the formal language. In a fifth step e), a manufacturing specification for manufacturing the product is determined using the formal skills description, the formal educt description and the formal product description, the manufacturing specification comprising a sequence of production steps, with each of the production steps one suitable for carrying out the production step Capability of at least one production facility is assigned.

This method has the advantage that a manufacturing specification can also be determined for products for whose manufacture previously unknown capabilities are required, provided that the formal relationships result in a continuous link from the control variables via the manipulated variables to the property parameters. Thus, the automation system is not limited to predefined capabilities of the production systems, but all conceivable capabilities can be derived from the formal capabilities description without an operator of the automation system having to make manual intervention.

In the present case, a capability of a production plant is understood to mean, in particular, that the production plant is able to carry out an action or manipulation specified by the capability. A capability has, in particular, a change of state as its objective, with the state of an educt preferably being changed during processing into the product. The change in state can, however, also relate to a production plant, for example exchanging a tool on a production plant and / or changing a parameter of a production plant can be referred to as a capability. The production facility can also be referred to as a production cell.

The automation system can also be referred to as an automated factory, robot factory or process plant. The automation system can be set up to manufacture discrete products, such as mechanical and / or electrotechnical components or products, and / or to manufacture formulated products, such as chemical and / or pharmaceutical products and / or foodstuffs. Educts are, for example, raw materials, blanks, individual components, semi-finished products and / or intermediate products that are suitable for further processing. A digital twin is understood to mean, for example, a digital image or digital model of a real existing entity, such as a production plant, an educt or a product. A digital twin can in particular have different levels of abstraction. For example, a digital twin can have a relatively high level of detail, that is to say a very detailed digital description of the respective entity, or also have a relatively low level of detail.

The digital twin of a production plant includes, for example, explicit functions, such as moving an object from a position A to a position B, and can also include more generally specified functions, such as an indication of all accessible positions (these can be defined explicitly or parametrically). One can speak of specific functions or skills and basic functions or skills. The digital twin can comprise a dynamic and / or a stationary model and / or a data-driven model of the production plant. A data-driven model is understood to mean, for example, that parameter values for a parameter equation that describes a specific relationship are determined on the basis of measurement data. This can also be called a Euchung or a calibration. A respective production plant can in particular comprise a number of production cells.

The digital twin of an educt includes, in particular, information on the characteristic property parameters of the educt. This includes, for example, information on the type of educt, on a composition of the educt, a quantity and / or geometric dimensions, as well as information on specific properties, in particular physical-chemical properties such as hardness, heat capacity, elasticity modulus, reactivity and the like, as well as biochemical Properties. The digital twin of the product includes, in particular, information on the characteristic property parameters of the product and on the predetermined product properties, for example as a target specification. This includes, for example, an indication of the type of product, a composition of the product, an indication of quantity and / or geometric dimensions. Furthermore, a predetermined product property in the form of a value of a physical measurement variable, such as a conductivity, a concentration and the like, can be contained in the digital twin.

Including that on the basis of the respective digital twin, a formal skill description, a formal educt description or a formal product description (in the following, in unambiguous cases, only a formal description can be used, whereby the reference results from the context) is generated, it is to be understood, for example, that a formal context that is contained in the respective formal description is read from the respective digital twin, that an information contained in the digital twin is generalized and / or abstracted on the basis an indication in the digital twin, a general context is determined and / or retrieved from a database or the like, and the like. The respective formal description can be generated automatically, for example by means of an algorithm or script, but it can also be generated by a user or operator of the automation system or a manufacturer of a production plant.

The formal description of the capabilities of the respective production plant includes a formal relationship, formulated in an abstract formal language, between a control variable of the production plant and a manipulated variable associated with the control variable. One can also say that the formal description represents a relationship between an input variable and an output variable of the production plant. As a control variable, for example, an input current or a Input voltage or a value of an input parameter. A variable that is dependent on the control variable can be considered as a manipulated variable, for example an amount of heat generated, an acceleration of a movement, forces that are caused by an acceleration, a generated luminous flux and the like. The formal context therefore depends on the type and, for example, a structure or set-up of the respective production facility. The formal context can also be referred to as a formal ability. For a conveyor belt system, for example, a relationship between an input parameter and different collection points can form a formal relationship, and a relationship between an input flow and a conveying speed can form a further formal relationship. For a heating plate, a relationship between the input current and the amount of heat generated can form the formal relationship. The formal description of the capabilities of a production plant is in particular more abstract than the digital twin of the production plant.

The formal relationship preferably includes a formally logical description, in particular abstractly by means of mathematical equations, functions and / or images, of a concrete relationship between the control variable and the manipulated variable. Depending on the complexity of the production system, several control variables and / or several manipulated variables can be contained in a formal context. A formal relationship can also be given in the form of a table of values, for example based on a measurement of the behavior of a manipulated variable as a function of the control variable or a property parameter as a function of the manipulated variable. The formal language corresponds to a syntax comprising characters, the meaning of which is clear. The formal language includes operators in particular to represent the relationships between the various variables, such as the control variable and manipulated variable, and the property parameters. One can also say that the formal language forms a semantic. The formal educt description includes, for example, a relationship between a manipulated variable of one of the production systems and one of the property parameters. Property parameters can be state variables or observables, for example. For example, a heating behavior of water in the form of an equation can represent such a formal relationship, with an amount of heat supplied as a manipulated variable and a specific heat capacity of water, an amount of water and a water temperature as a property parameter. Another formal context can describe the solubility (solution product) of a certain chemical compound in water, with the amount of heat supplied (indirectly via the temperature) also being the manipulated variable and the solubility being the property parameter.

The formal product description includes in particular a formal relationship between the predetermined product properties, such as a quantitative ratio of different substances, a temperature and / or a consistency, and the property parameters of the product, such as a composition, the temperature and / or a (statistical ) spatial distribution of components of the product.

The various formal descriptions and the formal relationships they contain can all be represented using the same formal language. Different formal descriptions can also contain formal relationships in different formal languages, provided that the linked variables, i.e. control variable, manipulated variable and / or property parameters, are identified in such a way that they can each be assigned to the corresponding variables in the other formal descriptions. For example, the temperature can appear in a first formal context as the capital letter “T”, in a second formal context as the word “temperature” and in a third formal context as the abbreviation “Temp.”. Therefore, from a synopsis of all formal descriptions, connections between them can be found capture. One can say that the formal relationships contained in the formal descriptions form a system of equations. In particular, the formal descriptions can be recorded by a computing unit and interfaces between different formal descriptions of different production systems, educts and / or products can be filtered or determined. An interface is, for example, the same manipulated variable or the same property parameter. Such interfaces can be determined, for example, by similarity algorithms, by a comparison of the type and / or the name and the like. It can be said that the formal descriptions of capabilities and formal educt descriptions of an automation system are integrated in a formal overall description. This formal overall description shows, for example, which of the production facilities of the automation system can interact with other components of the automation system by means of a mechanical force. On the basis of the formal descriptions, an arithmetic unit can therefore determine production steps completely independently of the predefined capabilities of a production plant.

Based on the formal descriptions, a manufacturing specification is therefore determined which comprises a sequence of production steps, the actual implementation of which leads to the product to be manufactured. The manufacturing specification is in particular automatically determined or derived, for example based on the formal overall description and the formal product description. The manufacturing specification can be referred to as a manual or a recipe. The manufacturing specification is not necessarily static, but a dynamic update, for example on the basis of current measurement data, can also be provided. Furthermore, the manufacturing specification can be continued and / or updated continuously, so that, for example, changing boundary conditions can be taken into account. Each of the production steps a capability of the at least one production plant that is suitable for carrying out the production step is assigned. Ability is understood here, for example, to mean that one of the formal abilities present in the formal skill description is used to achieve an effect or effect described by a formal context in the formal educt description. However, the capabilities can also include predetermined capabilities that are defined, for example, in the digital twin.

It should be noted that each production step that is contained in the manufacturing specification can itself be defined by means of a manufacturing specification. In this respect, complex products, which have to be assembled from a large number of semi-finished products, for example, can be represented by interleaved manufacturing instructions. In particular, different production systems can interact.

The determination of the manufacturing specification can in particular include calculating the system of equations formed by the formal relationships, for example by means of a control unit. For example, control variables can be varied in order to test or simulate the effect on the property parameters. In this respect, one can also speak of a simulated execution of the automation system. When determining the manufacturing specification, a trial-and-error method can also be used, in which different control variable values or sequences can be tried out with good luck. Different combinations of starting materials can also be tried out. In this way, it can be determined with a high degree of probability whether the production of the product with the automation system is possible in principle, or whether additional production systems and / or starting materials are necessary. A simple example involves mixing two liquids. The only production facility is a robotic arm with a gripper. There is no predefined ability "Mixing liquids". Taking into account the formal descriptions of the product and the educts, an arithmetic unit (formal logic) determines that a spatial distribution of the individual components must be achieved from being separate to being disordered (by mixing) in order to achieve the goal. Taking into account the formal descriptions of the production plant and the educts, the arithmetic unit (formal logic) determines that this goal can be achieved by "shaking" or by "stirring", with a container containing the two liquids as a whole for shaking is gripped by the gripper and then moved back and forth, and for stirring, the gripper is immersed in the container containing the two liquids and, for example, moved in a circular manner. The formal skills determined in this way are based entirely on the formal descriptions. In this example, there are thus two alternative manufacturing instructions.

The proposed method has the particular advantage that it is possible to produce constantly changing products without constant reprogramming of capabilities or re-parameterization of the production systems.

The method can also be used to design an automation system. Here, for example, the manufacture of different products is simulated on the basis of the digital twins and / or the different formal descriptions. This makes it possible to determine which skills are required or which skills could accelerate or make a manufacturing process cheaper.

According to one embodiment of the method, step e) further comprises determining skills required for manufacturing the product as a function of the formal skills Product description and the formal educt description, and checking a skills database which includes known skills for the at least one production plant of the automation system, depending on the skills required.

The required capabilities can in particular be determined independently of the production systems. For example, it is determined here that mixing of two liquids is possible by shaking (supplying kinetic energy into the liquid via the vessel) or by stirring (supplying kinetic energy into the liquid by a stirring rod).

Known capabilities include both predetermined capabilities, which are based, for example, on a specific function of the production system and are predetermined by a user, operator or manufacturer of the production system down to individual parameters, as well as formal capabilities of a production system, for example with an earlier production of a product. If a predetermined "stirring" ability is present, it is possible, for example, to dispense with determining a new "shaking" ability with a production plant. This embodiment is advantageous because it is possible to fall back on already known capabilities for manufacturing the product, as a result of which the computational effort for determining new capabilities based on the formal descriptions can be reduced.

According to a further embodiment of the method, this includes determining a new capability for the at least one production plant of the automation system on the basis of the formal capability description and the formal educt description.

The determination takes place in particular automatically by a computing unit. The new ability can be an ability which fulfills a purpose that has not yet been required, but can also represent an alternative to an already known ability. An example of this is mixing two liquids by stirring or shaking. All skills newly determined in this way form, in particular, formal skills.

According to a further embodiment of the method, this includes storing the determined new capability in the capability database for the at least one production facility of the automation system.

According to a further embodiment of the method, this includes optimizing the manufacturing specification as a function of a predetermined optimization parameter by assigning to each of the production steps of the manufacturing specification that capability of the at least one production plant that optimally performs the respective production step with respect to the optimization parameter.

Optimization parameters are, for example, a production time, the energy required for production, wear and tear caused by production, material scrap occurring during production, and the like.

In particular, new capabilities with optimized costs can be determined in a targeted manner in relation to an optimization parameter.

According to a further embodiment of the method, one of steps b), c) and / or d) comprises providing a physical model and / or a chemical model as a function of the digital twin for each production facility of the automation system, the digital twin for each of the starting materials and / or the digital twin of the product, and determining the formal skills description, the formal educt description and / or the formal Product description depending on the physical model and / or chemical model provided.

A physical model includes, for example, a physical relationship, such as Newton's second law, force = mass ^c acceleration. A chemical model includes, in particular, a chemical reaction equation, such as, for example, the dissolution of hydrogen chloride in water to form hydrochloric acid.

The physical model and / or the chemical model can in particular be provided by an external database and can be called up from this as required. On the basis of the physical model and / or the chemical model, formal relationships for the respective formal descriptions can be determined.

The physical model and / or the chemical model can in particular comprise a comparatively rough description of a real behavior of a physical or chemical variable or the course of a physical or chemical process.

This embodiment is advantageous when a respective digital twin contains only a small amount of information.

According to a further embodiment of the method, this includes simulating the physical model and / or the chemical model to determine the formal relationship between the control variable of the production plant and the associated manipulated variable, the formal relationship between the manipulated variable and the property parameters of the educts and / or the formal relationship between the predetermined product properties and the property parameters of the product.

This embodiment is advantageous because it allows completely new formal relationships to be determined. According to a further embodiment of the method, this comprises detecting a value of at least one of the property parameters of the at least one educt and / or a value of at least one of the property parameters of the product, and producing the product in accordance with the determined manufacturing specification and as a function of the detected value.

This embodiment is advantageous because the manufacture of the product is monitored. For example, different sensors can be provided in order to detect values of different property parameters.

According to a further embodiment of the method, this comprises producing a first product from the educts by means of a first production plant of the automation system according to a first production specification and producing a second product using the produced first product as an educt by means of a second production plant of the automation system according to a second manufacturing specification.

The first product can also be referred to as an intermediate product and the second product as the end product.

This embodiment has the advantage that for a complex product that is to be produced from the starting materials using a larger number of intermediate products, the entire production process does not have to be determined at once, but that the production of each of the intermediate products is determined separately and this then chained accordingly. In particular, it can be provided that for each of the production systems it is determined on the basis of the formal capabilities description whether it can produce a certain product (has the required capability) and which educts it requires for this. The required educts can then be seen as a product requirement, with each of the Product requirements like a product to be manufactured are considered. In this way, a production process for a complex product can be traced back to the existing starting materials as a chain of intermediate products.

According to a further embodiment of the method, the formal relationship includes an assignment of the control variable to the manipulated variable, an assignment of the manipulated variable to the property parameters of the educts and / or an assignment of the predetermined product property to the property parameters of the product in the form of an algebraic equation, a differential equation, a tensor equation, or a mapping.

In general, the formal relationship is, for example, a mapping from the space of the control variables into the space of the manipulated variables or a mapping from the space of the manipulated variables into the space of the property parameters or vice versa or a mapping from the space of the predetermined product properties into the space of the property parameters.

According to a further aspect, a computer program product is proposed which comprises instructions which, when the program is executed by a computer, cause the computer to carry out the proposed method.

A computer program product, such as a computer program means, for example, can be provided or delivered as a storage medium such as a memory card, USB stick, CD-ROM, DVD, or also in the form of a downloadable file from a server in a network. This can be done, for example, in a wireless communication network by transmitting a corresponding file with the computer program product or the computer program means.

According to a further aspect, an automation system with at least one production plant for the automated manufacture of a product with predetermined product properties is Shafts from a number of N starting materials (El - E3), with N> 1, proposed. The automation system comprises a control unit which is set up to carry out the proposed method.

The embodiments of the proposed method apply accordingly to the proposed automation system.

The control unit can be implemented in terms of hardware and / or also in terms of software. In the case of a hardware implementation, the control unit can be designed as a device or as part of a device, for example as a computer or as a microprocessor or as a control computer. In the case of a software implementation, the control unit can be designed as a computer program product, as a function, as a routine, as part of a program code or as an executable object

Further possible implementations of the invention also include not explicitly mentioned combinations of features or embodiments described above or below with regard to the exemplary embodiments. The person skilled in the art will also add individual aspects as improvements or additions to the respective basic form of the invention.

Further advantageous refinements and aspects of the inven tion are the subject matter of the subclaims and of the exemplary embodiments of the invention described below. In Wei direct the invention based on preferred embodiment forms with reference to the accompanying figures he explained in more detail.

Fig. 1 shows a schematic block diagram of an example for an automation system;

Fig. 2 shows a schematic block diagram of an example for determining a manufacturing specification based on formal descriptions; Fig. 3 shows a schematic block diagram of an example for an automation system; and

Fig. 4 shows a schematic block diagram of an example of a method for operating an automation system.

In the figures, elements that are the same or have the same function have been given the same reference symbols, unless otherwise stated.

1 shows a schematic block diagram of an example for an automation system 100. In the automation system 100, three production systems 110, 111, 112 are arranged by way of example. The automation system can also have more than three or fewer than three production systems 110, 111, 112. Production system 110 is configured, for example, as a robot arm with a gripper, production system 111 is configured, for example, as a climatic cabinet with an oven and a cooling compartment, and production system 112 is configured, for example, as a heating plate. Furthermore, by way of example, three educts E1, E2, E3 are in stock or available in automation system 100. For example, these are water El, olive oil E2 and an emulsifier E3. The automation system can also have more than three or less than three educts E1, E2, E3. The automation system 100 is to be used to manufacture the product P. The product P is characterized by predetermined product properties SPEC. For example, a skin cream P is to be produced.

A digital twin DU O, Dill, D112 is provided for each of the production systems 110, 111, 112, for example by the manufacturer of the respective production system 110, 111, 112. A digital twin is also created for each of the educts E1, E2, E3 DE1, DE2, DE3 provided, for example by the supplier of the respective educt El, E2, E3. Furthermore, a digital twin DP of the product P to be manufactured is provided. This can be done, for example, by automatically reading out the predetermined product properties SPEC, by an operator of the automation system 100 or by the customer who orders the product P.

A respective formal description F110, F111, F112, FEI, FE2, FE3, FP is generated on the basis of each digital twin DU O, Dill, D112, DE1, DE2, DE3, DP. This contains a formal relationship between a control variable XI and a manipulated variable X2 associated with it, between a property parameter X3 and a manipulated variable X2, or between a property parameter X3 and the predetermined product properties SPEC in a formal language. Thanks to the uniform use of the formal language, all formal relationships can be viewed together and mutual dependencies or possibilities for influencing can be determined.

A control unit 105, which, for example, forms a higher-level control entity for the automation system 100, links all formal descriptions F110, Fll, F112, FEI, FE2, FE3, FP in such a way that, for example, a coherent system of equations results. For example, a link table is created which contains all correlations between the existing control variables XI, manipulated variables X2, property parameters X3 and predetermined product properties SPEC. In this way, for example, a possibility can be found of how the product P can be produced from the existing educts E1, E2, E3 using the existing production systems 110, 111, 112.

To this end, the control unit 105 determines a production instruction PLAN, which here comprises, for example, a sequence of three production steps PA, PB, PC that are to be processed one after the other. Below is this for the example the skin cream P explained. The control unit 105 can, for example, use or employ optimization algorithms, genetic algorithms, machine learning methods, search algorithms, or algorithms from the field of satisfiability modulo theories (SMT) to determine the manufacturing specification PLAN.

The skin cream P to be produced consists, for example, of an emulsion with certain proportions of water E1 and olive oil E2, an emulsifier E3 being contained to stabilize the emulsion. An emulsion must therefore be prepared from the three components, but no "emulsification" ability is known. The control unit 105 must therefore itself determine how the emulsion can be produced on the basis of the formal descriptions F110, Fll, F112, FEI, FE2, FE3, FP.

For the robot arm 110, the digital twin DU O contains, for example, a maximum deflection for each axis of the robot arm 110, as well as maximum acceleration values for the movement of the arm in each of the axes. Based on this, the formal skills description F110 is generated, which includes, for example, a three-dimensional equation of motion as a formal relationship, with control variables XI being the motor currents for driving the respective axis and manipulating variables X2 a position, speed and acceleration of the gripper for each axis. Furthermore, the formal skills description F110 includes a formal context for the gripper, for example a control current as the control variable XI and a physical (gripping) force as the manipulated variable X2. For the climate cabinet 111, the digital twin Dill contains, for example, a range of values with regard to the maximum and minimum temperatures that can be reached in the oven and in the freezer compartment. From this, the formal skill description Fll is generated, which, for example, shows a formal relationship between an electrical output as the control variable XI and a heating output achieved as a result (for the Oven) or cooling capacity (for the freezer compartment) as the manipulated variable X2. Another formal relationship can describe, for example, a heating behavior over time (for the oven) or cooling behavior (for the ice compartment), with the heating output (for the oven) or the cooling output (for the freezer compartment) as the control variable XI and the temperature in linked to the respective area. For the heating plate 112, the digital twin D112 includes, for example, a function that describes a surface temperature as a function of an input parameter. The input parameter is, for example, translated into a control current by means of an assignment table that corresponds to a calibration. On the basis of the digital twin D112, the formal skills description F112 is generated for the heating plate 112, this for example including the input parameter as a control variable XI with the surface temperature as a manipulated variable X2 in the form of the function and as a second formal relationship Input parameters as control variable XI with the assignment table for the respective control current as manipulated variable X2.

The digital twins DE1, DE2, DE3 of the educts El, E2, E3 contain, for example, a chemical sum formula and an indication of the state of aggregation under normal conditions for the respective educt El, E2, E3. It can also contain information such as a package size. Based on this, the formal educt descriptions FEI, FE2,

FE3 is generated which, for example, includes values for physical quantities such as a heat capacity, a viscosity, a state diagram and the like in a formal context, for example a heating behavior or a flow behavior. Here, for example, a supplied amount of heat is linked as a manipulated variable X2 and the temperature as the property parameter X3 via the formal relationship. The formal relationships thus include in particular special physical equations that determine the behavior of the respective educt as a function of influencing variables and Describe environmental conditions. Furthermore, chemical reaction equations can exist as a formal relationship. The formal relationships can in particular be provided by an external unit, for example a server or a database.

The digital twin DP of the product P contains in particular the predetermined product properties SPEC. Based on this, the formal description FP is generated. This contains, for example, in the form of a formal relationship, an assignment of a property parameter X3, for example quantitative ratios of substances contained in the product P, to a value determined by the predetermined product properties SPEC.

The formal descriptions can be generated fully or partially automatically, but they can also be generated fully or partially manually. A formal description that has been generated once is preferably stored, for example together with the respective digital twin, so that it does not have to be continuously generated anew.

The formal descriptions F110, Fll, F112, FEI, FE2, FE3, FP are transferred to the control unit 105, which determines a manufacturing specification PLAN for manufacturing the product P from the educts E1, E2, E3 with the existing production systems 110, 111, 112 . The product P is an emulsion of water El and olive oil E2, which is stabilized by the emulsifier E3. Since the ability to "emulsify" or a production specification for producing an emulsion are not known, based on the formal educt descriptions FEI, FE2, FE3 and the formal product description FP, it is determined that the preparation of the emulsion must be carried out separately lying liquids must be mixed. Based on the formal skill descriptions F110, Fll, F112 and the formal educt descriptions FEI, FE2, FE3, it is determined that shaking a container containing the two liquids results in a thorough mixing of the two liquids. can be achieved and thus an emulsion can be produced. It is also determined that an increased temperature of the liquids is suitable for accelerating the process. The control unit 105 can thus determine a plurality of manufacturing regulations PLAN which differ in terms of individual production steps PA, PB, PC.

An exemplary production specification PLAN could include the following production steps PA, PB, PC: filling PA a vessel with water E1, olive oil E2 and emulsifier E3 according to their proportions in the emulsion by means of the robot arm 110; Heating PB the liquid mixture in the vessel to a predetermined temperature by means of the heating plate 112; and shaking PC of the closed vessel by means of the robot arm 110 until the emulsion is produced by the mixing of the liquids achieved. The emulsion can therefore be produced with the robot arm 110 and the heating plate 112, although an "emulsifying" ability is not known.

Fig. 2 shows a schematic block diagram of an example for determining a manufacturing specification PLAN on the basis of formal descriptions F110, FEI, FP. In this example, the formal relationships between a control variable XI, in this case a control current, a manipulated variable X2, in this case an amount of heat generated, and property parameters X3, in this case the temperature of educt El (see Fig. 1) and product P (see Fig. 1), given as an example.

The product P is hot water, the temperature being fixed at an interval in the form of the predetermined product properties SPEC. This temperature interval is given in the formal product description FP in the form of two inequalities, where the actual temperature T of the product P represents the property parameter X3 and the predetermined specification SPEC by a lower limit Tmin (e.g. 80 ° C) and a upper limit Tmax (e.g. 105 ° C) is included. The educt El is, for example, water. The formal educt description FEI contains, as a formal relationship, an equation which sets the temperature T as the property parameter X3 in relation to the supplied amount of heat Q as the manipulated variable X2. Here, TO is the starting temperature of the water El, c is the specific heat capacity of water and m is the mass of the water to be heated El.

The production facility 110 (see FIG. 1) is, for example, a heating plate. The formal capability description F110 of the heating plate 110 contains an equation as a formal relationship which represents a generated amount of heat Q as the manipulated variable X2 as a function of the control current I (t), which is variable over time, as the controlled variable XI. In this case, U denotes a defined control voltage, tO is a starting point in time and tl is an end point in time for energizing the heating plate 110.

The control unit 105 relates the formal relationships to one another, that is, it links the same variables with one another. In the present case, the control unit 105 recognizes in particular that the water El can be heated by the heating plate 110, the manipulated variable X2 (the amount of heat Q) corresponding to one another in the formal context of the heating plate 110 and the water El. For example, the control unit 105 carries out a simulation, which is shown in FIG. 2 by two diagrams. The left diagram shows a time profile of the control current XI for the heating plate 110. At a point in time t0, the control current I is set to the maximum value Imax, which leads to a heat supply to the water El and thus to an increase in temperature. The right diagram shows the time course of the temperature X3 of the water El. The two limit values Tmin, Tmax known from the formal product description FP are also entered here. The temperature X3 of the water El it increases starting from the starting value T0, for example linearly, as long as the control current XI is constant. At time tl, the temperature X3 is above the lower limit value Tmin, which is why the control current XI is switched off. The product P is thus produced from the educt E1.

Based on this simulation, the control unit 105 determines the manufacturing specification PLAN, which includes a single production step PA, which causes the heating of the water El.

Fig. 3 shows a schematic block diagram of an example for an automation system 100. The automation system 100 comprises three production systems 110, 111,

112 controlled by the control unit 105.

In this example, the control unit 105 is connected to a first database 101, which stores the digital twins DU O, Dill, D112 of the production systems 110, 111, 112 arranged in the automation system 100 and makes them available on request. Furthermore, the control unit 105 is connected to a second database 102 which stores the digital twins DE1, DE2, DE3 of the educts E1, E2, E3 (see FIG. 1) present in the automation system 100 and makes them available on request. Formal descriptions Fl10, Fl11, Fl12, FEI, FE2, FE3 are generated based on the digital twins DU O, Dil, D112, DE1, DE2, DE3. The formal descriptions F110, Fll,

F112, FEI, FE2, FE3 are contained in the digital twins DU O, Dil, D112, DE1, DE2, DE3 or are stored together with them in the respective database 101, 102.

An order for a product P (see FIG. 1) is fed to the control unit 105 in the form of a digital twin DP of the product P. The control unit 105 first determines whether all necessary educts E1, E2, E3 are available for the automatic production of the product P and whether the necessary skills are known. To this end, the control unit 105 queries a connected skills database 120. The capabilities database 120 stores, for example, all of the capabilities of the production systems 110, 111, 112 before certain capabilities and, in addition, all of the formats newly determined over time. len skills. If all the required skills are known in the skills database 120, the control unit 105 determines a manufacturing specification PLAN for the product P from the known skills. Otherwise, the control unit 105 tries, as described above with reference to FIGS. 1 or 2, based on the formal descriptions F110, Fll, F112, FEI, FE2, FE3, FP to determine a new formal capability in order to automatically manufacture the product P to make possible with the available means.

Fig. 4 shows a schematic block diagram of an example of a method for operating an automation system 100 (see Fig. 1 or 3). The automation system 100 comprises at least one production plant 110,

111, 112 (see Fig. 1 or 3) for the automated production of a product P (see Fig. 1) with predetermined product properties SPEC (see Fig. 1 or 2) from a number of N starting materials E1, E2, E3 (see Fig . 1), where N> 1. The product P and each educt El, E2, E3 can be characterized by a number of property parameters X3 (see Fig. 1 or 2).

In a first step S1, a digital twin DU O, Dill, D112 (see Fig. 1 or 3) for each production plant 110, 111, 112 of the automation system 100, a digital twin DE1, DE2, DE3 (see Fig. 1 or 3 ) for each of the N educts E1, E2, E3 and a digital twin DP (see Fig. 1 or 3) of the product P to be produced is provided.

In a second step S2, a formal skill description F110, Fll, F112 (see FIGS. 1 - 3) is created for each production plant 110, 111, 112 on the basis of the respective digital twin DU O, Dill, D112 of the production plant 110, 111, 112 is generated, the formal ability description F110, Fll, F112 a formal relationship between a control variable XI (see FIG. 1 or 2) of the production plant 110, 111, 112 and one of the control variable XI associated manipulated variable X2 (see Fig. 1 or 2) in a formal language.

In a third step S3, a formal educt description FEI, FE2, FE3 (see Fig. 1 - 3) for each of the N educts El, E2, E3 on the basis of the digital twin DE1, DE2, DE3 of the respective educt El, E2, E3 generated, the respective formal educt description FEI, FE2, FE3 comprising a formal relationship between the manipulated variable X2 and the property parameters X3 of the educt El, E2, E3 in the formal language.

In a fourth step S4, a formal product description FP (see FIGS. 1-3) of the product P is generated on the basis of the digital twin DP of the product P, the formal product description FP showing a formal relationship between the predetermined product properties SPEC and the property parameters X3 of the product (P) in the formal language.

In a fifth step S5, a manufacturing specification PLAN (see FIGS. 1 - 3) for manufacturing the product P using the formal skill description F110, Fll, F112, the formal educt description FEI, FE2, FE3 and the formal product Description FP determined, the manufacturing regulation PLAN comprising a sequence of production steps PA, PB, PC (see Fig. 1 or 2), each of the production steps PA, PB, PC being one suitable for carrying out the production step PA, PB, PC Capability of the at least one production plant 110, 111, 112 is assigned.

Further steps of the method can include an optimization of a production instruction PLAN and / or a production step PA, PB, PC as a function of an optimization parameter. For example, an increased temperature could be used for the shortest possible production time, but this would mean a higher energy consumption. The method can also be used to simulate an automation system 100. For example, by means of such a simulation, an optimal composition of the automation system 100 with regard to the equipment with production systems 110, 111, 112 can be determined. Furthermore, an extension of an existing automation system 100 can be simulated with an additional production plant, on the basis of which a better purchase decision for the operator of the automation system 100 is possible.

Although the present invention has been described using exemplary embodiments, it can be modified in many ways. It can thus be provided that each individual production plant is viewed as an automation system and is operated as described above. Each production plant can thus manufacture products from educts independently of the other production plants. If a complex product is to be manufactured, a chain of production steps carried out by different production plants may be necessary. The production plants in the association can, for example, “order” an educt that corresponds to a product from another production plant. For example, a production plant for making a soup requires hot water, which it requests (orders) from a production plant for heating liquids, which in turn requests the requested amount of water from a bottling plant for water. In this way, the association of production systems can automatically produce complex products in a very flexible manner. List of reference symbols

100 automation system

101 database

102 database

105 control unit

110 production plant

111 production facility

112 production plant 120 database

DE1 digital twin DE2 digital twin DE3 digital twin DU O digital twin Dill digital twin D112 digital twin DP digital twin El educt E2 educt E3 educt

F110 formal description Fll formal description F112 formal description FEI formal description FE2 formal description FE3 formal description FP formal description P product PA production step PB production step PC production step PLAN manufacturing specification

51 procedural step

52 Process step

53 procedural step

54 procedural step

55 Process step SPEC predetermined product properties XI control variable X2 manipulated variable X3 property parameters

Claims

1. A method for operating an automation system (100) which has at least one production plant (110-112) for the automated production of a product (P) with predetermined product properties (SPEC) from a number of N starting materials (El-E3), with N> 1 , wherein the product (P) and each educt (El - E3) can be characterized by a number of property parameters (X3), the method comprising: a) providing (Sl) a digital twin (DU O - D112) for each production plant (110 - 112) of the automation system (100), a digital twin (DE1 - DE3) for each of the N starting materials (El - E3) and a digital twin (DP) of the product to be manufactured (P), b) generation (S2) a formal skill description (F110 - F112) for each production plant (110 - 112) based on the respective digital twin (DU O - D112) of the production plant (110 - 112), the formal skill description (F110 - F112) being one formal connection between e A control variable (XI) of the production plant (110 - 112) and a manipulated variable (X2) associated with the control variable (XI) in a formal language, c) generating (S3) a formal educt description (FEI - FE3) for each of the N educts (El - E3) based on the digital twin (DE1 - DE3) of the respective educt (El - E3), whereby the respective formal educt description (FEI - FE3) shows a formal relationship between the manipulated variable (X2) and the property parameters (X3 ) of the educt (El - E3) in the formal language, d) generating (S4) a formal product description (FP) of the product (P) on the basis of the digital twin (DP) of the product (P), the formal product -Description (FP) comprises a formal relationship between the predetermined product properties (SPEC) and the property parameters (X3) of the product (P) in the formal language, and e) determining (S5) a manufacturing specification (PLAN) for manufacturing the product (P) using the formal Skills description (F110 - F112), the formal educt description (FEI - FE3) and the formal product description (FP), the manufacturing specification (PLAN) comprising a sequence of production steps (PA, PB, PC), whereby each of the production steps (PA, PB, PC) is assigned a capability of the at least one production plant (110-112) that is suitable for carrying out the production step (PA, PB, PC).

2. The method according to claim 1, characterized in that step e) further comprises:

Determination of skills required to manufacture the product (P) as a function of the formal product description (FP) and the formal educt description (FEI - FE3), and checking a skills database (120), which known skills for the at least one Production plant (110-112) of the automation system (100), depending on the required capabilities determined.

3. The method according to claim 1 or 2, characterized by: determining a new capability for the at least one production plant (110-112) of the automation system (100) on the basis of the formal skills description (F110 - F112) and the formal educt description ( FEI - FE3).

4. The method according to claim 3, characterized by: storing the determined new capability in the capabilities database (120) for the at least one production plant (110-112) of the automation system (100).

5. The method according to any one of claims 1 to 4, characterized by:

Optimizing the manufacturing specification (PLAN) as a function of a predetermined optimization parameter in that each of the production steps (PA, PB, PC) of the manufacturing specification (PLAN) is assigned the capability of the at least one production system (110) - (112) that corresponds to the respective Performs the production step (PA, PB, PC) optimally in relation to the optimization parameters.

6. The method according to any one of claims 1 to 5, characterized in that one of steps b), c) and / or d) further comprises:

Provision of a physical model and / or a chemical model depending on the digital twin (DU O - D112) for each production plant (110 - 112) of the automation system (100), the digital twin (DE1 - DE3) for each of the educts (El - E3) and / or the digital twin (DP) of the product (P), and

Determination of the formal skills description (F110 - F112), the formal educt description (FEI - FE3) and / or the formal product description (FP) depending on the physical model and / or chemical model provided.

7. The method according to claim 6, characterized by: simulating the physical model and / or the chemical model to determine the formal relationship between the control variable (XI) of the production plant (110-112) and the associated manipulated variable (X2), the formal relationship between the manipulated variable (X2) and the property parameters (X3) of the educts (El - E3) and / or the formal relationship between the predetermined product properties (SPEC) and the property parameters (X3) of the product (P).

8. The method according to any one of claims 1 - 7, characterized by:

Detecting a value of at least one of the property parameters (X3) of the at least one educt (El-E3) and / or a value of at least one of the property parameters (X3) of the product (P), and

Manufacture of the product (P) according to the determined manufacturing specification (PLAN) and depending on the recorded value.

9. The method according to any one of claims 1 - 8, characterized by:

Production of a first product (P) from the educts (El-E3) by means of a first production plant (110-112) of the automation system (100) in accordance with a first production specification (PLAN), and

Production of a second product (P) using the produced first product (P) as an educt (El-E3) by means of a second production plant (110-112) of the automation system (100) according to a second production specification (PLAN).

10. The method according to any one of claims 1 - 9, characterized in that the formal relationship is an assignment of the control variable (XI) to the manipulated variable (X2), an assignment of the manipulated variable (X2) to the property parameters (X3) of the educts (El - E3) and / or an assignment of the predetermined product property (SPEC) to the property parameters (X3) of the product (P) in the form of an algebraic equation, a differential equation, a tensor equation, and / or a mapping.

11. Computer program product, comprising instructions which, when the program is executed by a computer, cause the computer to execute the method according to any one of claims 1-10.

12. Automation system (100) with at least one production plant (110-112) for the automated production of a product (P) with predetermined product properties (SPEC) from at least one educt (El-E3), with N> 1, and with a control unit (105 ), which is set up to carry out the method according to any one of claims 1-10.