A METHOD AND APPARATUS FOR DYNAMICALLY OPTIMIZING INDUSTRIAL PRODUCTION PROCESSES

The invention relates to a method and apparatus for optimiz ing dynamically an industrial production process of a produc tion plant consisting of a plurality of physical production modules .

A production plant which can produce a wide variety of dif ferent products can comprise a plurality of physical produc tion modules, in particular manufacturing machines and/or transport facilities to transport workpieces during the pro duction process. The industrial production process comprises sequences of production process steps. The production modules of the production plant can be subject to changes such as up grades, production line changes, product changes and/or changes due to repair and/or maintenance. Whenever a physical production module is altered or exchanged, an associated digital twin data model can be changed accordingly. This data model change of a digital twin data model of the physical module can be either performed manually or automatically. For instance, an engineer can update the digital twin data model of a physical module which has been upgraded. However, the change of a production module can lead to unwanted, wrong or sub-optimal process sequences of the production process.

Critical production module changes such as a replacement of a whole production module can require a complete process recal culation by an ERP system or even manually by human opera- tors . In a conventional industrial production plant, when it comes to critical production module changes production lines or process segments must first be shutdown, before the respec tive physical production modules of the production plant can be altered or replaced. Further, the respective software com ponents, i.e. the data models of the production modules must be changed. Finally, the production line process segment can be restarted and tested together with the respective changed or updated software components. This conventional approach is very time-consuming and involves extensive manual activities both for physical installations and/or software-related changes .

Accordingly, it is an object of the present invention to pro vide a method and apparatus for optimizing dynamically an in dustrial production process wherein required installation and adaptation times are significantly reduced.

This object is achieved according to a first aspect of the present invention by a dynamic process optimizer apparatus comprising the features of claim 1.

The invention provides according to the first aspect a proc ess optimizer apparatus for optimizing dynamically an indus trial production process of a production plant including physical production modules,

wherein said process optimizer comprises:

a watchdog component adapted to monitor the production mod ules of said production plant to detect configuration changes within said production plant,

a model comparator component adapted to evaluate a production plant data model of the production plant comprising digital twin data models related to physical production modules of said production plant to identify automatically deviating model elements of digital twin data models related to physi cal production modules of said production plant affected by the configuration changes detected by said watchdog component and

a process resequencer component adapted to perform a dynamic process optimization of at least one production process of said production plant depending on the deviating model ele ments identified by said model comparator component.

In a possible embodiment of the process optimizer apparatus according to the first aspect of the present invention, the watchdog component of said process optimizer apparatus is connected to a communication infrastructure of the production plant to detect configuration changes within the production plant comprising hardware configuration changes and/or soft ware configuration changes related to physical production modules of the production plant.

In a further possible embodiment of the process optimizer ap paratus according to the first aspect of the present inven tion, the watchdog component is fed with plant data of said production plant via a data interface.

In a still further possible embodiment of the process opti mizer apparatus according to the first aspect of the present invention, the production plant data model comprises digital twin data models of the physical production modules of said production plant stored in a local memory of said process optimizer apparatus or stored in a remote database connected to said process optimizer apparatus. In a still further possible embodiment of the process opti mizer apparatus according to the first aspect of the present invention, the production plant data model comprises as twin data models for the physical production modules of said pro duction plant attribute-value lists indicating capabilities of the respective production modules.

In a further possible embodiment of the process optimizer ap paratus according to the first aspect of the present inven tion, the model comparator component is implemented to per form rule-based attribute-to-attribute comparisons in the at tribute-value lists of the stored production plant data model for production modules affected by configuration changes de tected by said watchdog component to identify automatically deviating model elements, in particular value changes in the attribute-value lists.

In a further possible embodiment of the process optimizer ap paratus according to the first aspect of the present inven tion, the production plant data model comprises a semantic data model representing an ontology comprising digital twin data models characterizing capabilities of the physical pro duction modules and dependencies between the physical produc tion modules of said production plant.

In a further possible embodiment of the process optimizer ap paratus according to the first aspect of the present inven tion, the process resequencer component is adapted to deter mine if the deviating model elements identified by said model comparator component have effects and/or implications on pro duction process sequences of production process steps per formed by production modules of the production plant and is adapted to provide, if an effect has been identified, a dy- namic resequencing of production process sequences performed by production modules of said production plant based on a predefined set of relations specifying dependencies between physical production modules and production process steps of the production process.

In a possible embodiment of the process optimizer apparatus according to the first aspect of the present invention, the production process steps of the production process sequences are stored in an ERP database.

In a still further possible embodiment of the process opti mizer apparatus according to the first aspect of the present invention, the process optimizer apparatus is adapted to per form the automatic optimization of the industrial production process of said production plant during runtime of the pro duction plant on the basis of real-time data received by the watchdog component of the process optimizer apparatus moni toring the production modules of said production plant.

In a still further possible embodiment of the process opti mizer apparatus according to the first aspect of the present invention, the process optimizer apparatus is adapted to per form the automatic optimization of the industrial production process of said production plant during a simulation session where an industrial production process of said production plant is simulated.

In a still further possible embodiment of the process opti mizer apparatus according to the first aspect of the present invention, the production modules of said production plant comprise transport production modules adapted to transport workpieces and/or intermediate products between predefined positions to perform associated production process steps, wherein said transport production modules comprise in par ticular robot arms and/or conveyor belts.

In a still further possible embodiment of the process opti mizer apparatus according to the first aspect of the present invention, the digital twin data model of a physical produc tion module is updated automatically when the corresponding physical production module is upgraded.

In a still further possible embodiment of the process opti mizer apparatus according to the first aspect of the present invention, the process optimizer apparatus is implemented on a local server of the production plant.

In a still further possible alternative embodiment of the process optimizer apparatus according to the first aspect of the present invention, the process optimizer apparatus is im plemented on a remote server connected to the communication infrastructure of the production plant.

The invention further provides according to a further aspect a production plant comprising the features of claim 14.

The invention provides according to the second aspect a pro duction plant comprising a plurality of production modules adapted to perform production process steps of a production process and comprising a process optimizer apparatus accord ing to the first aspect of the present invention.

The invention further provides according to a third aspect a method for optimizing dynamically an industrial production process of a production plant including physical production modules ,

the method comprising the steps of:

monitoring the production modules of said production plant to detect configuration changes within said production plant, evaluating a production plant data model of the production plant comprising digital twin data models related to physical production modules of the production plant to identify auto matically deviating model elements of digital twin data mod els related to physical production modules of said production plant affected by the detected configuration changes, performing a dynamic process optimization of at least one production process of said production plant depending on the identified deviating model elements.

In the following, possible embodiments of the different as pects of the present invention are described in more detail with reference to the enclosed figures.

Fig. 1 shows a schematic diagram of a system comprising a process optimizer apparatus according to the first aspect of the present invention;

Fig. 2 shows a flowchart of a possible exemplary embodi ment of a method for optimizing automatically an industrial production process according to the sec ond aspect of the present invention.

In the schematic diagram of Fig. 1, a dynamic process opti mizer (DPO) apparatus 1 according to the first aspect of the present invention is provided for optimizing dynamically an industrial production process of a production or manufactur ing plant including different physical production modules. The production plant 2 can comprise a plurality of different and complex physical production modules 3-i to perform proc ess steps. Physical production modules 3-i can comprise ma chines performing process steps on work pieces and/or inter mediate products during the production process. The produc tion modules 3-i can also comprise physical transport produc tion modules adapted to transport workpieces and/or interme diate products between predefined positions to perform asso ciated production process steps. These transport production modules can comprise for instance robot arms of robot units and/or conveyor belts. In the illustrated example of Fig. 1, the production plant 2 comprises a shop floor with different physical production modules 3-1, 3-2 ... 3-N as illustrated in Fig. 1. The physical production modules can be connected to each other in series and/or in parallel to provide a network of production modules which are adapted to perform production process sequences consisting of one or more process steps.

The number of production modules 3-i implemented in the shop floor of the production plant 2 can vary depending on the complexity of the industrial production process. In a possi ble embodiment, the dynamic process optimizer apparatus 1 comprises an interface and is connected to the communication infrastructure of the production plant 2. The communication infrastructure of the production plant 2 can comprise a local area network LAN to which the different physical production modules 3-i, in particular machines and/or transport elements are connected. The connection to the communication infra structure can comprise a wireless and/or a wired connection. In a possible embodiment, the process optimizer apparatus 1 comprises three main components as also illustrated in Fig.

1. In the illustrated embodiment, the dynamic process optimizer (DPO) apparatus 1 comprises a watchdog component 1A, a model comparator component IB and a process resequencer component 1C. The watchdog component 1A of the process optimizer appa ratus 1 is adapted to monitor the production modules 3-i of the production plant 2 to detect configuration changes within the monitored production plant 2. The model comparator compo nent IB of the process optimizer apparatus 1 is configured or adapted to evaluate a production plant data model or a twin data model of the production plant 2 which comprises digital twin data models related to physical production modules 3-i of the production plant 2 to identify automatically deviating model elements of digital twin data models related to physi cal production modules of the production plant 2 affected by the configuration changes detected by the watchdog component 1A. The production plant data model PPDM (digital twins) can be stored in a local memory and/or a remote database 4 as il lustrated in Fig. 1. The production plant data model PPDM comprises a number of digital twin data models DMs related to corresponding physical production modules 3-i of the produc tion plant 2. In a possible embodiment, the production plant data model PPDM comprises as twin data models DMs for the physical production modules 3-i of the production plant 2 at tribute-value lists indicating capabilities of the respective production modules 3-i. The capabilities of each production module 3-i in the production plant 2 can be formulated in a possible embodiment in simple attribute-value lists. For in stance, a production module 3-1 formed by a robot arm can comprise for the attribute number of axes the value Num- berAxis Ax=3. In a possible embodiment, the model comparator component IB can be implemented to perform a rule-based dis tributed to attribute comparisons in the attribute-value lists of the stored production plant data models for produc- tion modules 3-i affected by configuration changes detected by the watchdog component 1A to identify automatically devi ating model elements, in particular value changes in the at tribute-value lists, as a model comparison result MCR. Direct attribute-to-attribute comparisons can be performed by the model comparator component IB of the process optimizer appa ratus 1 based on a rule-based system. For instance, if a ro bot arm as a physical production module 3-1 is exchanged by a replacing robot arm 3R-1 capabilities and/or characteristics of the physical production module 3-1 can be changed. For in stance, the new robot arm replacing the previous robot arm may comprise a higher number of production axes. For in stance, the previous production plant module 3-1 formed by a robot arm with only three axes (Ax=3) can for instance be re placed by a new robot arm forming a new replacing production module 3R-1 having four axes (Ax=4) . In this example, the performed configuration change within the production plant 2 is a hardware configuration change where an existing produc tion module 3-i such as a robot arm is replaced by a replace ment production module 3R-i. This hardware configuration change can be detected by the monitoring watchdog component 1A of the dynamic process optimizer apparatus 1. The watchdog component 1A is adapted to detect automatically configuration changes performed at the production plant 2, in particular hardware configuration changes when hardware components are changed (e.g. upgraded) or replaced. In a further possible embodiment, the watchdog component 1A of the dynamic process optimizer 1 can also be adapted to detect software configura tion changes related to physical production modules 3-i of the production plant 2. For instance, a local control soft ware component can be reconfigured or updated causing a soft ware configuration change observed by the watchdog component 1A of the dynamic process optimizer apparatus 1. The model comparator component IB is adapted to automatically identify deviating model elements of digital twin data models DMs re lated to physical production modules 3-i of the production plant 2 affected by the configuration changes detected by the watchdog component 1A. For instance, if a robot arm 3-1 hav ing three axes (Ax=3) is replaced by a robot arm 3R-1 having four axes (Ax=4), an attribute in the corresponding attrib ute-value list forming a digital twin data model of the re spective production module changes its stored value from the previous value 3 to the new value 4. In a possible embodi ment, the digital twin data model of a physical production module is updated automatically when a corresponding physical production module 3-i of the production plant 2 is upgraded or changed. In a possible embodiment, the memory or database 4 where the production plant data model is stored can com prise an interface with the communication infrastructure of the shop floor of the production plant 2. In this embodiment, the different digital twin data models DMs of the production modules 3-i can be automatically updated, automatically when the corresponding physical production module 3-i is upgraded physically .

The dynamic process optimizer (DPO) apparatus 1 further com prises a process resequencer component 1C which is adapted to perform a dynamic process optimization of at least one pro duction process of the production plant 2 depending on the deviating model elements identified by said model comparator component IB. In a possible embodiment, the process rese quencer component 1C is configured to determine if the devi ating model elements identified by said model comparator com ponent IB as a model comparison result MCR have effects and/or implications on production process sequences of pro duction process steps performed by production modules 3-i of the production plant 2. If an effect or implication has been identified by the process resequencer component 1C, a dynamic resequencing of production process sequences performed by the production modules of the production plant 2 can be performed by the process resequencer component 1C based on a predefined set of relations which specify dependencies between physical production modules 3-i and product process steps of the pro duction process. In a possible embodiment, the production process steps of the production process sequences are stored in an ERP database 5 as illustrated in Fig. 1. In the illus trated embodiment, the industrial process comprises a number X of process steps PS to be performed by the production plant 2. In a possible embodiment, the process resequencer compo nent 1C can comprise a rule engine which works in two phases. In a first phase, the process resequencer component 1C finds out if the deviating model elements received from the model comparator component IB do have an effect on the currently queued production/process steps. For instance, the first process step PS1 can consist of substeps SSI-1, SSI-2 and SS1-X as also illustrated in Fig. 1. In the production plant 2, substep SSI-1 can be performed by production module 3-1 which can comprise a robot arm that lifts a workpiece onto a conveyor belt 3-N along a trajectory or moving route. The model comparator component IB can detect that production mod ule 3-1 was exchanged by a replacement production module 3R-1 offering improved capabilities because of an additional axis. This improved capability can have as a consequence or effect that substep SSI-2 of the production process is no longer re quired. The detection of implications and/or effects due to model changes currently ongoing and/or future plant produc tion sequences for specific products can be performed in the first phase by the process resequencer component 1C based on a predefined set of relations. Each relation can specify de- pendencies between a physical production module 3-i and pro duction steps PS available in the production plant 2. For in stance, if the first production module 3-1 is a robot arm having only three axes (Ax=3) , a motion pattern could be that a workpiece is moved from a beginning position A to an inter mediate positon B and then from the intermediate position B to the final position C. From there, it is moved to D. The intermediate position B might be necessary because the robot arm has only three axes (Ax=3) .

In contrast, after the robot arm has been replaced by the up graded robot arm 3R-1 having four axes (Ax=4), the moving pattern can be that the workpiece is directly moved from the starting position A to the final position C. In this phase, the process substep SSI-2 becomes obsolete. The set of rela tions can be formulated based on a system's data model, e.g. a relational database, a semantic data model and/or a rule- based system. The execution of process replanning and/or re sequencing in the second phase can be realized by invoking a default production planning mechanism of the specific produc tion plant 2. The process resequencer component 1C of the dy namic process optimizer apparatus 1 can provide a dynamic process resequencing based on a predefined set of relations. In a possible embodiment, the replanned production steps can be re-entered in a plant's default production execution sys tem.

As illustrated in Fig. 1, an initial process logic PLog of production module 3-1 is substituted by an improved process logic IPLog for the replaced production module 3R-1.

In a possible embodiment, the production plant data model PPDM comprises a semantic data model representing an ontology comprising digital twin data models DMs which characterize capabilities of the physical production modules and which characterize also dependencies between the different physical production modules 3-i of the production plant 2. In a possi ble embodiment, the production plant data model PPDM can com prise an OWL semantic data model. The semantic modelling al lows to detect subtle and complex data dependencies. For in stance, the capabilities of a new production module can be inferred to be similar with the capabilities of the previous replaced production module, although single attributes dif fer. This can be inferred due to possible replacements of one attribute for another. For example, for some production steps, it might for instance not matter if a robot arm offers four axes instead of three axes. It is possible that the stored ontology is traversed by a query language QL such as SPARQL .

In a possible embodiment, the dynamic process optimizer (DPO) apparatus 1 comprising the watchdog component 1A can receive sensor data from sensors placed in the production plant 2 to monitor different production modules 3-i of the production plant 2 and to receive real time or machine data from the production plant 2. In a possible embodiment, the dynamic process optimizer apparatus 1 is adapted to perform the auto matic optimization of the industrial production process of the production plant 2 during runtime of the production plant 2 on the basis of the received real-time data. In an alterna tive embodiment, the dynamic process optimizer apparatus 1 can also be configured to perform the automatic optimization of the industrial production process in a simulation session where the industrial production process of the production plant 2 is simulated. In a possible implementation, the dy- namic process optimizer apparatus 1 can operate in a real time operation mode and/or in a simulation mode.

The process optimizer apparatus 1 can be implemented in a possible embodiment on a local server of the production plant 2. In an alternative embodiment, the process optimizer appa ratus 1 is implemented on a remote server connected to the communication infrastructure of the production plant 2, for instance via a data network. The dynamic process optimizer apparatus 1 has access to the memory or database 4 where the production plant data model PPDM is stored and to a memory or database 5 including the ERP information of the industrial process. In a possible embodiment, the improved process logic IPLog calculated by the dynamic process optimizer apparatus 1 for one or more production modules can be rewritten into the ERP memory 5 for further use.

Fig. 2 shows a flowchart of a possible exemplary embodiment of a method for optimizing dynamically an industrial produc tion process of a production plant 2.

In a first step SI, the production modules 3-i of the produc tion plant 2 are monitored to detect a configuration change within the production plant 2. This monitoring step can for instance be performed by a watchdog component 1A of a process optimizer apparatus 1. The detected configuration changes can comprise hardware configuration changes and/or software con figuration changes related to the physical production modules 3-i of the production plant 2.

In a further step S2, a memorized production plant data model of the production plant 2 comprising digital twin data models DMs related to physical production modules 3-i of the produc- tion plant 2 is evaluated to identify automatically deviating model elements of digital twin data models related to physi cal production modules 3-i of the production plant 2 affected by the configuration changes detected in step SI. The evalua tion of the production plant data model PPDM in step S2 can for instance be performed by a model comparator component IB of a process optimizer apparatus 1.

In a further step S3, a dynamic process optimization of at least one production process of the production plant 2 is performed depending on the deviating model elements identi fied in step S2. The dynamic process optimization can be per formed in a possible embodiment by a process resequencer com ponent 1C of a process optimizer apparatus 1. The method il lustrated in Fig. 2 can be performed in real time during op eration of the production plant 2 or in a simulation session during a planning phase of the industrial production plant 2. With the method and system 1 according to the present inven tion, installation and adaption time of data models and/or software models relevant to a production plant creation can be greatly reduced. Further, upgrades of production modules 3-i within an existing production plant 2 are facilitated. Moreover, it is possible to avoid in many cases a shutdown of the production plant 2 when changing or replacing parts and/or physical production modules of the production plant 2.