Computer optimization of engineered modular device topology

Abstract

A computer-implemented method (100) for optimizing a given topology (2) of a modular plant (1) that is to perform a given industrial process according to a given engineering recipe (3), the method (100) comprising: • for at least one process module (21-23) in the given topology (2), obtaining (110) an amount of at least one resource and / or performance (21a-23a) of the process module (21-23) that is utilized when the process is performed according to the recipe (3), and dividing the amount by a maximum amount of the respective resource and / or performance (21a-23a) that the process module (21-23) is capable of providing, thereby obtaining a theoretical utilization (21b-23b) of the resource and / or performance (21a-23a) as a theoretical utilization of the process module (21-23); • searching (120) in a pool (6) of available process modules (24-27) for candidate process modules (24-27) that are capable of replacing at least one process module (21-23) in the execution of the given recipe (3) and that fit the given topology (2); • for each candidate process module (24-27), obtaining (130) a theoretical utilization (24b-27b) of a corresponding resource and / or performance (24a-27a) of the candidate process module (24-27) that would ensue if the at least one process module (21-23) were to be replaced by the candidate process module (24-27), and assigning the theoretical utilization (24b-27b) to the candidate process module (24-27); and • generating an optimized topology (2) of the plant (1) from the given topology (2) by replacing at least one process module (21-23) by a candidate process module (24-27) that has the same or a higher theoretical utilization (24b-27b) than the at least one process module (21-23) * ).

Description

TECHNICAL FIELD

[0001] The present invention relates to the optimization of a topology of a modular industrial plant that is to perform a given industrial process according to a given recipe and that is assembled from reusable modules. BACKGROUND

[0002] In many industrial applications, there is a need to quickly reconfigure a plant from one production process to another. For example, the product being produced can be evolving quickly and each improvement in quality should be brought into the production process as soon as possible to maximize customer satisfaction. There are also products whose end effect per unit of quality or volume is so concentrated that the supply for a year can only take a few weeks to produce. A prime example of such products are pharmaceutical compositions that are delivered to patients in doses on the order of tens or hundreds of milligrams.

[0003] The plant for manufacturing such compositions can be assembled from independent modules that receive one or more educts on one or more input ports, process the one or more educts into one or more products by performing one or more physical and / or chemical actions, and deliver the one or more products to one or more output ports. During the production of a particular compound, the modules are linked together in an ad hoc manner. When the production of this compound is complete and the next compound needs to be produced, the modules are separated from each other so that they can be reused in a different configuration. EP 3 318 935 Al discloses an exemplary method of operating physical process modules in a modular process plant.

[0004] Currently, creating a design plan for a modular plant that performs a given industrial process is a human engineering task.

[0005] Invention objectives

[0006] It is an object of the present invention to at least partially automate the engineering process by using a computer to transform an engineered topology that can perform a given process into an optimized topology that can perform the same process but better utilizes the physical process modules.

[0007] This object is achieved by a computer-implemented method for optimizing a given topology of a modular plant and by a method for manufacturing a modular industrial plant. SUMMARY

[0008] The inventors have developed a computer-implemented method for optimizing a given topology of a modular industrial plant to be executed for a given industrial process. The industrial process as a whole can take in one or more educts and transform these educts into one or more products by executing a series of specific tasks given in a recipe. These tasks are executed by process modules assembled according to the topology. The method starts from a given topology, which can be a topology designed by a human, but also a topology that has been optimized in one or more computerized steps starting from such a human-designed topology.

[0009] The method starts by obtaining for at least one process module in the given topology a quantity of at least one resource and / or performance that the module utilizes when the process is executed according to the recipe. This quantity is divided by the maximum quantity of the respective resource and / or performance that the process module can provide. This yields a theoretical utilization of the resource and / or performance, which can also be seen as a theoretical utilization of the process module.

[0010] A pool of available modules is searched for candidate process modules that can replace the at least one process module in the execution of the given recipe and that fit into the given topology. In other words, if the at least one process module under consideration is replaced by one of the candidate process modules, the given process is still executed according to the same given recipe.

[0011] For each candidate process module, a theoretical utilization of the corresponding resource and / or performance of the candidate process module is obtained in case the at least one process module under consideration is replaced by this candidate process module, and the theoretical utilization is assigned to the candidate module. From the given topology, an optimized topology is generated by replacing the at least one process module currently under consideration by a candidate process module that has the same or preferably a higher theoretical utilization than the at least one process module.

[0012] If the theoretical utilization of the new module in the optimized topology is higher than the theoretical utilization of the old module in its place in the previous topology, this indicates that the quantity of resources and / or performance now installed in that place more closely matches the demand for that resource and / or performance.

[0013] If the theoretical utilization of the new module is the same as the theoretical utilization of the old module, the new module is at least equivalent to the old module in terms of resource utilization. Among such equivalent modules, one module can be chosen according to any other optimization criterion. In this way, the topology of the plant can be improved with respect to this other criterion without necessarily becoming worse in terms of resource utilization.

[0014] The inventors have found that this partial automation reduces the cost of the plant both in terms of physical process modules and in terms of human engineering time, thereby increasing the productivity of the final industrial plant assembled from process modules according to the topology.

[0015] Starting from a recipe of an industrial process to be executed, an engineer can easily come up with at least one topology that allows to execute the process. For example, the engineer can choose to select most of the modules in the topology from a small set of generic modules, each providing a lot of different process functionality. This minimizes engineering time, but can not be optimal for the outcome of the process and for the productivity of the site where the several modular devices assembled from the set of generic available modules will be operated.

[0016] There is a general tendency that the more different process functionalities a single module combines, the worse the performance of the module is when executing each of the single processes, as the combination of all these functionalities requires some technical trade-off. Therefore, for each process functionality, one can expect that a process module dedicated to that specific process functionality has the best performance. This is somewhat analogous to the computing domain: a general purpose CPU can be able to execute any task, but if one wants to execute a task in the fastest and most energy efficient way (like mining bitcoins), there is no way around an ASIC dedicated to this task.

[0017] At the site where the modular industrial devices are operated, usually more than one such device is operated at any given time. The concurrently operated devices are usually assembled from process modules taken from a common pool of process modules. This introduces interdependencies between the topologies of the concurrently operated devices, and the efficiency of the utilization of the modules needs to be evaluated on a site-wide basis, i.e. across the boundaries between individual devices.

[0018] In a toy example, process A is to be executed on a first device and requires a module that is able to heat a substance, process B is to be executed on a second device and requires a module that is able to heat and stir a substance, and the pool of available modules includes a heater, a stirrer, and one generic module that is able to heat and stir. While using the generic module for process A is a viable solution, it has the disadvantage that process B requires two additional modules (a heater and a stirrer), so that together processes A and B require three modules. It would be better for process A to use the simple heater and for process B to use the generic module; this saves the use of the stirrer.

[0019] Therefore, by improving the utilization of individual process modules, the topology of a device to execute one process can be made more "friendly" to other processes to be run on different devices at the same site, without any need to know anything about these other processes. The optimization happens only within the scope of one process, which is computationally easier to handle than the optimization of all devices at a given site.

[0020] The standard of theoretical utilization introduces an abstraction layer behind which any complexity of the process can be "hidden". When considering a particular process module, optimization is focused only on that process module without any other information about what happens to the product of that process module later. Thus, theoretical utilization is a very useful framework that can condense the optimization of a topology (a technique previously performed by engineers) into a numerical optimization problem that can be easily solved by a computer. Specifically, a large set of candidate process modules can be created that can replace the process module under consideration, and from this large set of candidate process modules, the best candidate process module can be found by a computer. Thus, the optimization of a topology is not only taken from the field of human engineering and automated "as is". Rather, it is transformed into a form that is easily handled by a computer but is more difficult to understand by an engineer.

[0021] A given topology can specifically include

[0022] • at least one input of the industrial process as a feed to at least one input port of the process module;

[0023] • at least one output of the industrial process as a delivery from at least one output port of the process module; and

[0024] • at least one interconnection between an output port of a first process module and an input port of a second process module.

[0025] Typically, one input of the process (i.e., an educt) will be passed through multiple process modules as a whole and interact with at least one other such input before finally being converted as a whole into an output of the process (i.e., a product).

[0026] An engineering recipe can specifically include a time sequence of process steps. Each step can include sending at least one command to at least one process module and / or performing at least one action by at least one process module. Based on the recipe, it is known at any time which process functions are needed, so a process module topology that is capable of performing the process according to the recipe can be designed directly. Furthermore, any subsequent modification of the topology proposed by the optimization can be easily checked to determine whether the topology is still compatible with the recipe after the modification.

[0027] In a recipe, at least one transition between consecutive steps can specifically include waiting for at least one process module to reach a state that satisfies at least one predetermined criterion. This simplifies the engineering of any other steps, as the above-mentioned state can be considered a given state.

[0028] In a particularly advantageous embodiment, for each available process module, the pool of available process modules comprises a mapping between the services provided by the process module and the process functions from a predetermined list. In this way, the terminology regarding the services can be unified between different modules. For example, the service "tempering" of a first process module, the service "heating" of a second process module, the service "tempering" of a third process module can all provide the process function of "tempering".

[0029] For each available process module, the pool of available process modules can also comprise a mapping between the service parameters and the limits of these service parameters and a mapping between the function parameters of the process functions and the limits of these function parameters. In this way, the parameters and their limits can also be unified. For example, a first process module capable of tempering can specify a maximum voltage of a heating coil, a second process module capable of tempering can specify a maximum current through the heating coil, and a third process module capable of tempering can specify a maximum wattage of the heating coil. All these different specifications can be mapped to a respective maximum temperature that can be used during tempering.

[0030] In another particularly advantageous embodiment, the resources and / or performance whose utilization is evaluated specifically comprise a count of the respective services of the at least one process module or candidate process module under consideration. This high-level abstraction allows to convert the utilization into a number that can be used for optimization without any need to understand details about the internal processes of the modules.

[0031] For example, the service can specifically comprise one or more of the following:

[0032] • heating or cooling a substance and / or maintaining the temperature of a substance at a desired value;

[0033] • stirring a substance;

[0034] • filling at least one container with a desired amount of a substance;

[0035] • discharging a desired amount of a substance from at least one container;

[0036] • adding a desired amount of a second substance to a first substance;

[0037] • mixing a mixture of two or more substances by mechanical interaction with the mixture of the two or more substances;

[0038] • distilling at least one substance from a mixture of two or more substances;

[0039] • converting at least one substance; and

[0040] • inerting at least one substance.

[0041] In another particularly advantageous embodiment, the resources and / or performance specifically include an amount of at least one educt processed per unit of time, and / or an amount of at least one product produced per unit of time. Many process modules have different quantitative performance, and the performance used should ideally match the performance actually used. For example, if the batch size is 1 cubic meter, it is not optimal to use a stirring module that can hold 3 cubic meters of mixture. This stirring module can need to be in another device at the same site, where the batch size is actually 3 cubic meters; if it is tied to a place where only 1 cubic meter, another 3 cubic meter module needs to be purchased at additional expense. Furthermore, a 3 cubic meter module takes up more valuable floor space than actually needed in a 1 cubic meter process. If heating is needed during the stirring process, the volume inside the heating module is 3 cubic meters instead of 1 cubic meter, which can also consume too much energy.

[0042] In another particularly advantageous embodiment, the resources and / or performance specifically include a measurement range and / or a dynamic range of the respective at least one measuring instrument of the at least one process module or candidate process module. Similar to the previous example, if a measuring instrument with a large measurement range is tied to a process that uses only a small part of this range, it can not be used in another device at the same site that needs a large measurement range. Furthermore, if only a small part of the measurement range is used, the resolution of the measurement can be affected. In many measuring instruments, an analog-to-digital converter maps the analog measurement value from the measurement range to a digital output value represented in a fixed number of bits (e.g., 32 bits). If only one-eighth of the measurement range is actually used, the information is encoded in one-eighth of the range of the digital output value, i.e., only 4 bits. In this case, most of the information acquired is discarded. Only in the case that the measurement range is used completely, the measurement result has full resolution.

[0043] In another particularly advantageous embodiment, determining whether a candidate process module from the pool of available modules is suitable for a given topology specifically includes:

[0044] • for each input port of the at least one process module utilized according to the given topology, determining whether the candidate process module has a corresponding input port;

[0045] • for each output port of the at least one process module utilized according to the given topology, determining whether the candidate process module has a corresponding output port; and

[0046] • if the candidate process module has a corresponding port for each utilized input port and output port of the at least one process module, determining that the candidate process module is suitable for the given topology.

[0047] This ensures that each connection that should be made to the process module under consideration according to the topology can also be made to the candidate process module.

[0048] The checking whether the candidate process module is suitable for the given topology can optionally be postponed until the candidate process module enters the "shortlist" of candidate process modules, which can bring it into the optimized topology by virtue of its good theoretical utilization. This can save calculation time during the optimization, in particular in case the checking whether the candidate process module is suitable for the given topology takes longer than the calculation of the theoretical utilization of the candidate process module.

[0049] In the context of this checking, the corresponding input port of the candidate process module can in particular be an input port that provides the same process functionality as the input of the at least one process module under consideration. Likewise, the corresponding output port of the candidate process module can in particular be an output port that has the same process functionality as the output port of the at least one process module. In this way, it can be more accurately determined whether the candidate process module is suitable for the given topology, in particular in case the candidate process module is a generic module that combines different process functionalities in one physical module. If an input port or an output port of the candidate process module is free, but the port belongs to a different process functionality than the process functionality used in the at least one process module to be replaced by the candidate process module, the process will be executed differently after the change and can violate the given recipe.

[0050] For example, if the process module under consideration has two output ports for a metering service and the candidate process module also has two output ports for a metering service, this is compatible with the given topology. But if the candidate process module has only one output port for a metering service, while its second output port is for a conditioning service, this is not compatible with the given topology: if the candidate process module replaces the process module under consideration, the input port of another process module connected to the second output port of the candidate process module should receive the substance processed by the metering service, but receives the conditioned substance instead.

[0051] Determining whether the candidate process module is suitable for the given topology can further comprise:

[0052] • determining whether all ports of the candidate process module that are needed for the utilized services of the candidate process module to run are connected in a state in which all connections to the ports of the candidate process module have been made according to the given topology; and

[0053] • in response to determining that the utilized services of the candidate process module cannot run due to a loss of a connection to a port, determining that the candidate process module is not suitable for the given topology.

[0054] For example, the old process module under consideration can only need connections for the input and output of a substance to work, but the new generation candidate process module can also need a signal connection to a safety interlock system to stop the process in case of a detected leak of the substance. If this candidate process module replaces the old process module, it will not work, because the safety interlock is not connected.

[0055] In another particularly advantageous embodiment, in response to determining that two or more candidate process modules have the same theoretical utilization rate equal to or higher than the theoretical utilization rate of the at least one process module, the candidate process module with the lowest total number of input and output ports is selected as the candidate process module to replace the at least one process module. The total number of input and output ports is another indicator of the utilization rate of a candidate process module: if two candidate modules can perform the same role in the industrial process according to the recipe, but one of them has extra ports, those extra ports do not necessarily play a role in the industrial process as described above. Like the theoretical utilization rate discussed above, this indicator is an abstract indicator that does not require detailed knowledge of the internal processes of the candidate process module.

[0056] For example, two candidate process modules can only provide conditioning services used in the context of the process, so that in terms of services they both have a full theoretical utilization rate. However, while the first candidate process module only has an input for the substance to be conditioned and an output for the conditioned substance, the second candidate process module also has output ports for various measurements collected during conditioning. Since both candidate process modules can be used to replace the process module under consideration, the output ports for the measurements do not seem necessary for conditioning in the context of the process. Therefore, the first candidate process module is preferably used. The second candidate process module can be needed in another device elsewhere in the site where conditioning under supervision using the measurements is required.

[0057] In another particularly advantageous embodiment, for two or more candidate process modules having a theoretical utilization rate at least as high as the theoretical utilization rate of the at least one process module, the value of at least one predetermined key performance indicator is evaluated that would occur in turn if the at least one process module were to be replaced by the candidate process module. The candidate process module with the best result value of the key performance indicator is selected as the candidate process module to replace the at least one process module. The key performance indicator is a comprehensive quality factor, for example, which can be attributed to the modular industrial plant, the process performed by the plant, the combination of multiple plants or processes running at the same site, or even the entire site where multiple plants or processes are run. Thus, the key performance indicator can be used as a tool to evaluate the improvement brought by the optimization of the topology across plant boundaries.

[0058] In this way, the optimization can at least partly concentrate on predetermined key performance indicators. A better value of the key performance indicator results in the candidate process module becoming preferred under the boundary condition that the theoretical utilization at least does not worsen.

[0059] The key performance indicator can in particular comprise one or more of: a cost of executing the industrial process; a total number of modules in the industrial plant; a throughput of one or more products from one or more educts of the industrial process as a whole; and an energy consumption of the industrial process. In particular, the cost of executing the industrial process can comprise a cost of use of the modules. For example, there can be multiple specialized candidate modules that can execute a particular process function and improve the theoretical utilization because they do exactly what is needed and nothing more. But on the other hand, using a highly specialized super-high-quality module can be more expensive than if the module used was less specialized than the module previously envisaged but more general than a highly specialized module (and thus used less).

[0060] In another particularly advantageous embodiment, for a plurality of different process modules in a given topology, the amount of utilization of at least one resource / performance is in particular acquired. From each of these different process modules, a respective candidate optimized topology of the plant is generated. For each candidate optimized topology, a value of at least one key performance indicator that would ensue if the candidate optimized topology were to be implemented is determined. From the values of the key performance indicators and at least one optimality criterion, one optimized topology is selected from the candidate optimized topologies.

[0061] The reason behind this is that there is typically a limited physical process module inventory at the site where the modular plant is to be run. There can be multiple places in a given topology where a candidate process module can be substituted for another process module to improve utilization and / or a key performance indicator. In particular, there can be more such places than available instances of the candidate process module. This means that a decision has to be made where to place the available instances of the candidate process module for maximum effect.

[0062] For example, there can be two places in a given topology where only a single simple stirring function is needed. At a first place, the topology contains a multi-functional process module that has three functions in addition to stirring. At a second place, the topology contains a more multi-functional process module that has five functions in addition to stirring. If a simple stirring module is inserted at the second place, a more valuable module can be freed up for use at another place in the topology, or even in another plant at the same site.

[0063] In another particularly advantageous embodiment, the amount of utilization of at least one resource and / or performance is specifically obtained for a combination of two or more interconnected process modules. The search is specifically performed for candidate process modules that are able to replace the combination of process modules and that are suitable for the given topology. The optimized topology of the plant is generated by replacing the combination of two or more process modules by a candidate process module that has the same or a higher theoretical utilization than the combination of two or more process modules.

[0064] In this way, multiple functions located in interconnected modules can be merged into a candidate process module without creating new underutilization. For example, a combination of a mixing module and a conditioning module can be replaced by a combined mixing conditioning module that does not provide other services. However, a multi-function module that provides five other services in addition to heating and mixing would not be proposed as a candidate process module, because the five other services would be idle. Merging multiple functions into one process module can save floor space and energy, but this advantage can be overcompensated if a multi-function module is used for this purpose without actually using all of the module.

[0065] The optimized topology can be presented to the engineer in any suitable way. For example, when rendering a given topology on a display, any process module that the optimized topology suggests to replace can be rendered differently from other modules to indicate that a beneficial replacement is available. For example, the process module can be rendered in a light color or a different color, or a box can be drawn around the process module in the rendering. The process module can be made clickable so that clicking on it causes the display to show suggested replacements from a pool of available modules.

[0066] The invention also provides a method for manufacturing a modular industrial plant to perform a given industrial process according to a given engineering recipe. According to the method, starting from a given engineering topology of the plant, an optimized topology of the plant is generated using the above-described computer-implemented method. The physical process modules are then physically connected according to the optimized topology, thereby producing the modular industrial plant.

[0067] As mentioned above, many of the advantages of these methods are brought about by the computerization of the methods. Therefore, the invention also provides a computer program having machine-readable instructions that, when executed by one or more computers and / or industrial control systems, cause the one or more computers and / or industrial control systems to perform one of the above-described methods. The invention also provides a non-transitory computer storage medium and / or a download product having the computer program. BRIEF DESCRIPTION OF DRAWINGS

[0068] In the following, the invention will be explained using the accompanying drawings, which are not intended to limit the scope of the invention. The drawings show:

[0069] Figure 1 : exemplary embodiment of a method 100;

[0070] Figure 2 : example of an engineering topology 2 with process modules 21-23 which can be optimized using a pool 6 of available modules 24-27;

[0071] Figure 3 : exemplary embodiment of a method 200 for manufacturing a modular plant 1. DETAILED DESCRIPTION

[0072] Figure 1 is a schematic flow chart of an exemplary embodiment of a method 100 for optimizing a given topology 2 of a modular industrial plant 1. The modular plant 1 is to perform a given industrial process according to a given engineering recipe 3. The topology 2 can be designed to perform this recipe and / or it can also be a result of a previous optimization.

[0073] In step 110, for at least one process module 21-23 of the given topology 2, an amount of at least one resource and / or performance 21a-23a of the process module 21-23 is obtained which is utilized when the process is performed according to the recipe 3. This amount is divided by a maximum amount of the respective resource and / or performance 21a-23a which the process module 21-23 is able to provide. The result is a theoretical utilization 21b-23b of the resource and / or performance 21a-23a which can also be regarded as a utilization of the process module 21-23. It is pre-set which resource and / or performance 21a-23a is to be investigated.

[0074] In step 120, a candidate process module 24-27 is searched in the pool 6 of available process modules 24-27 which is able to replace the considered process module 21-23 and which is also suitable for the given topology 2.

[0075] In particular, for each input port I of the at least one process module 21-23 which is utilized according to the given topology 2, it can be determined in block 121 whether the candidate process module 24-27 has a corresponding input port I. Then, for each output port O of the at least one process module 21-23 which is utilized according to the given topology, it can also be determined in block 122 whether the candidate process module 24-27 has a corresponding output port O. If both determinations are positive (truth value 1), it can be determined in block 123 that the candidate process module 24-27 is suitable for the given topology 2.

[0076] Alternatively or in combination, it can be determined in block 124 whether all ports I, O of the candidate process module 24-27 are connected which are required for the utilized services of the candidate process module 24-27 to run in a state in which all connections to the ports I, O of the candidate process module 24-27 have been made according to the given topology 2. If it is determined in a subsequent block 125 that the utilized services of the candidate process module 24-27 cannot be run due to a loss of connection to the ports I, O (true value 1), it can be determined in a block 126 that the candidate process module 24-27 is not suitable for the given topology 2.

[0077] In a step 130, for each candidate process module 24-27, the theoretical utilization 24b-27b of the corresponding resources and / or performance 24a-27a of the candidate process module 24-27 is obtained which would occur in the event that at least one process module 21-23 would be replaced by this candidate process module 24-27. This theoretical utilization 24b-27b is assigned to the candidate process module 24-27. Here, the resources 24a-27a whose theoretical utilization 24b-27b is determined correspond to the resources 21a-21a of the process module 21-23 under consideration.

[0078] In a step 140, an optimized topology 2 of the device 1 is generated from the given topology 2 by replacing at least one process module 21-23 by a candidate process module 24-27 which has the same or a higher theoretical utilization 24b-27b than the at least one process module 21-23 * . As mentioned before, this process can be repeated. That is, the optimized topology 2 * may be fed again to the step 110 to increase the utilization in more places. For example, a module 21-23 which has been utilized according to the topology 2 before can be put into the pool 6 of available modules after it has been released in the optimized topology 2 * to replace a module which has occupied the place in the optimized topology 2 * . This can open up new possibilities for further increasing the utilization.

[0079] According to block 141, it can be determined whether two or more candidate process modules 24-27 have the same theoretical utilization 24b-27b which is equal to or higher than the theoretical utilization 21b-23b of the at least one process module 21-23. If this is the case (true value 1), according to block 142, the candidate process module 24-27 can be selected which has the smallest total number of input ports I and output ports O as the candidate process module 24-27 to replace the at least one process module 21-23.

[0080] According to block 143, for two or more candidate process modules 24-27 having a theoretical utilization 24b-27b at least as high as the theoretical utilization 21b-23b of at least one process module 21-23, the value of the at least one predetermined key performance indicator 7 that would occur in case the at least one process module 21-23 would be replaced by the candidate process module 24-27 is evaluated. Then, according to block 144, the candidate process module 24-27 having the best result value of the key performance indicator 7 is selected as the candidate process module 24-27 to replace the at least one process module 21-23. Which value of the key performance indicator 7 is best can be determined according to any suitable optimality criterion 8.

[0081] According to block 115, the amount of utilization of the at least one resource and / or performance 21a-23a can be specifically acquired for a plurality of different process modules 21-23 in the given topology 2. After steps 120 and 130 as described before, then, in block 145 within step 140, a respective candidate optimized topology 2' for optimization of each different process module 21-23 device 1 from which optimization starts can be generated. In step 150, the value of the at least one key performance indicator 7 that would occur in case the candidate optimized topology (2') would be implemented can be determined. In step 160, according to the value of the key performance indicator 7 and the at least one optimality criterion 8, one optimized topology 2 * may be selected from the set of candidate optimized topologies 2'.

[0082] According to block 117, the amount of utilization of the at least one resource and / or performance 21a-23a can be specifically acquired for a combination of two or more interconnected process modules 21-23 instead of only for one process module 21-23. Then, according to block 127, the search of step 120 can be specifically performed for candidate process modules 24-27 that are able to replace the combination of process modules 21-23 instead of only one process module 21-23. The candidate process modules 24-27 still need to fit into the given topology 2 as would be the case if they only replace one process module 21-23. According to block 147 in step 140, the optimized topology 2 * of the device can be generated by replacing the combination of two or more interconnected process modules 21-23 by a candidate process module 24-27 having the same or a higher theoretical utilization 24b-27b than the combination of the two or more process modules 21-23 instead of only one process module 21-23. The way the theoretical utilization is defined for the combination of modules is exactly the same as for a single module.

[0083] Figure 2A simple example of an engineering topology 2 of a modular industrial plant 1 is shown. The topology 2 comprises a conditioning module 21, a combined mixing-agitation module 22, and a distillation module 23. The input I * The input I is fed as a whole to the input port I of the conditioning module 21. The output O of the conditioning leaves the conditioning module 21 and enters the combined mixing-agitation module 22 at the input port I of this module. The agitated substance leaves the output port O of the mixing-agitation module 22 and enters the distillation module 23 at the input port I. The output O of the distillation is the output O of the process to be run as a whole * and leaves the distillation module 23 at the output port O.

[0084] The pool 6 of available modules 24-27 comprises one agitation module 24 with two input ports I and one output port O, one agitation module 25 with one input port I and one output port O, a combined agitation-conditioning module 26, and another distillation module 27.

[0085] In the engineering topology 2, the conditioning module 21 and the distillation module 23 are fully utilized. They each provide one service as a resource 21a, 23a, and this service is required for the process to be executed, so the utilization 21b, 23b is 1. Therefore, these modules 21b, 23b that can improve utilization have no replacement. However, the distillation module 23 can be replaced by the distillation module 27, which will then have the same utilization 27b of 1. This can be advantageous if the distillation module 27 is advantageous in some key performance indicators 7 compared to the distillation module 23.

[0086] The combined mixing-agitation module 22 provides two services, i.e. agitation and mixing, as a resource 22a. According to the topology only one service, i.e. agitation, is used, so the utilization 22b of the combined module 22 is only 0.5.

[0087] Replacing the mixing-agitation module 22 by the agitation module 24 will result in a utilization 24b of 1. However, in this example the agitation module 24 has two input ports I and it is desired that both are connected for agitation. Therefore, while the agitation module 24 will improve utilization, it is not suitable for the topology 2 because it will not work if inserted instead of the combined module 22.

[0088] The agitation module 25 has one input port I and one output port O. This means that the agitation module 25 is suitable for the topology 2. Its utilization 25b is also 1 when inserted instead of the combined module 22. Therefore, replacing the combined module 22 by the agitation module 25 improves utilization and frees the combined mixing-agitation module 22 for other uses of other equipment at the same site.

[0089] The conditioning function of conditioning module 21 and the stirring function used by combined module 22 can also be combined into one module by replacing modules 21 and 22 with combined conditioning and stirring module 26. Module 26 provides two services as its resource 26a, and resource 26a will have a utilization rate 26b of 1. However, module 26 has two output ports O, and it is expected that both stirring and conditioning are connected if both are required. Therefore, similar to module 24, combined module 26 cannot replace modules 21 and 22 because it is not suitable for topology 2.

[0090] Figure 3 This is a schematic flowchart of an exemplary embodiment of a method 200 for manufacturing modular industrial equipment 1 to perform a given industrial process according to a given engineering formula 3. In step 210, starting from a given engineering topology 2 of equipment 1, an optimized topology 2 of equipment 1 is generated using the previously described computer-implemented method 100. * In step 220, based on optimized topology 2 * Physical connection optimization topology 2 * It includes physical process modules 21-27. In this way, modular industrial equipment 1 is produced.

[0091] List of reference numerals

[0092] 1: Modular industrial equipment

[0093] 2: Given topology for device 1

[0094] 2 * Optimize topology

[0095] 2': Candidate Optimization Topology

[0096] 21-23: Process Modules in Topology 2

[0097] 21a-23a: Resources for modules 21-23

[0098] 21b-23b: Theoretical utilization rate of resources 21a-23a

[0099] 24-27: Available process modules in pool 6

[0100] 24a-27a: Resources for modules 24-27

[0101] 24b-27b: Theoretical utilization rate of resources 24a-27a

[0102] 3: The given formula to be executed by device 1

[0103] 6: Pools available for process modules 24-27

[0104] 7: Key Performance Indicators

[0105] 8: optimal criterion

[0106] 100: method for optimizing a given topology 2

[0107] 110: obtaining theoretical utilization rates 21b-23b

[0108] 115: considering different process modules 21-23 as starting points

[0109] 117: considering combinations of modules 21-23

[0110] 120: searching for suitable modules 24-27 in pool 6

[0111] 121: determining matching of input ports I of modules 24-27

[0112] 122: determining matching of output ports O of modules 24-27

[0113] 123: determining that candidate modules 24-27 are suitable for topology 2

[0114] 124: determining whether all required ports are connected

[0115] 125: determining whether missing connections impede service

[0116] 126: determining that candidate modules 24-27 are not suitable for topology 2

[0117] 127: considering combinations of modules 21-23

[0118] 130: obtaining theoretical utilization rates 24b-27b

[0119] 140: generating optimized topologies 2 *

[0120] 141: determining that multiple modules 24-27 have the same utilization rate

[0121] 142: considering total number of input ports I and output ports O

[0122] 143: evaluating values of key performance indicators 7 of modules 24-27

[0123] 144: selecting candidate modules 24-27 with best performance indicators 7

[0124] 145: obtaining multiple candidate optimized topologies 2'

[0125] 147: considering combinations of modules 21-23

[0126] 150: determining key performance indicators 7 of candidate topologies 2'

[0127] 160: Select optimized topology 2 from candidate 2' set *

[0128] 200: Method for manufacturing modular industrial plant 1

[0129] 210: Generate optimized topology 2 by method 100 *

[0130] 220: Physically connect optimized topology 2 * Modules 21-27 in

Claims

1. A method (100) for a modular industrial plant (1) that is to perform a given industrial process according to a given engineering recipe (3), the method (100) comprising: • for at least one process module (21-23) in a given topology (2) of the modular industrial plant (1), obtaining (110) a quantity of at least one resource and / or performance (21a-23a) of the process module (21-23) that is utilized when the process is performed according to the engineering recipe (3), and dividing the quantity by a maximum quantity of the respective resource and / or performance (21a-23a) that the process module (21-23) is capable of providing, thereby obtaining a theoretical utilization (21b-23b) of the resource and / or performance (21a-23a) as a theoretical utilization of the process module (21-23); • searching (120) in a pool (6) of available process modules (24-27) for candidate process modules (24-27) that are capable of replacing the at least one process module (21-23) in the performance of the engineering recipe (3) and that fit the given topology (2); • for each candidate process module (24-27), obtaining (130) the theoretical utilization (24b-27b) of the corresponding resource and / or performance (24a-27a) of the candidate process module (24-27) that would ensue if the at least one process module (21-23) were to be replaced by the candidate process module (24-27), and assigning the theoretical utilization (24b-27b) to the candidate process module (24-27); • generating an optimized topology (2) of the modular industrial plant (1) from the given topology (2) by replacing the at least one process module (21-23) by a candidate process module (24-27) having the same or a higher theoretical utilization (24b-27b) compared to the at least one process module (21-23) ); receiving one or more educts at one or more input ports of the modular industrial plant (1); processing the one or more educts into one or more products by the modular industrial plant (1) by performing the engineering recipe (3); and delivering the one or more products to one or more output ports of the modular industrial plant (1).

2. The method (100) of claim 1, wherein the topology (2) specifically comprises: at least one input (I ) of the industrial process as a whole to at least one input port (I) of the process module (21-23); at least one output (O ) of the industrial process as a whole from at least one output port (O) of the process module (21-23); and at least one interconnection between an output port (O) of a first process module (21-23) and an input port (I) of a second process module (21-23).

3. The method (100) of any one of claims 1-2, wherein the engineering recipe (3) specifically comprises a time sequence of process steps, wherein each step comprises sending at least one command to at least one process module (21-23) and / or performing at least one action by at least one process module (21-23).

4. The method (100) of claim 3, wherein at least one transition between consecutive steps in the engineering recipe (3) specifically comprises waiting for at least one process module (21-23) to reach a state that satisfies at least one predetermined criterion.

5. The method (100) of any one of claims 1-2, wherein for each available process module (24-27), the pool (6) of available process modules (24-27) comprises a mapping between a service provided by the process module (24-27) and a process function from a predetermined list.

6. The method (100) according to claim 5, wherein for each available process module (24-27), the pool (6) of available process modules (24-27) further comprises, on the one hand, a mapping between service parameters and limits of the service parameters and, on the other hand, a mapping between functional parameters of the process function and limits of the functional parameters.

7. The method (100) according to any one of claims 1 to 2, wherein the resources and / or capabilities (21a-27a) specifically comprise a count of respective services of the at least one process module (21-23) or of the candidate process module (24-27).

8. The method (100) according to claim 7, wherein the services specifically comprise one or more of: • heating or cooling a substance and / or maintaining a temperature of the substance at a desired value; • stirring a substance; • filling at least one container with a desired amount of a substance; • discharging a desired amount of a substance from at least one container; • adding a desired amount of a second substance to a first substance; • mixing a mixture of two or more substances by mechanical interaction with the mixture; • distilling at least one substance from a mixture of two or more substances; • converting at least one substance; and • inerting at least one substance.

9. The method (100) according to any one of claims 1 to 2, wherein the resources and / or capabilities (21a-27a) specifically comprise: an amount of at least one educt processed per unit of time and / or an amount of at least one product produced per unit of time.

10. The method (100) according to any one of claims 1 to 2, wherein the resources and / or capabilities (21a-27a) specifically comprise: a measurement range and / or a dynamic range of at least one measuring instrument of the at least one process module (21-23) or of the candidate process module (24-27).

11. The method (100) according to any one of claims 1 to 2, wherein determining whether a candidate process module (24-27) from the pool (6) of available modules (24-27) is suitable for the given topology (2) specifically comprises: • for each input port (I) of the at least one process module (21-23) utilized according to the given topology (2), determining (121) whether the candidate process module (24-27) has a corresponding input port (I); • for each output port (O) of the at least one process module (21-23) utilized according to the given topology, determining (122) whether the candidate process module (24-27) has a corresponding output port (O); and • if the candidate process module (24-27) has a corresponding port (I, O) for each utilized input port (I) and output port (O) of the at least one process module (21-23), determining (123) that the candidate process module (24-27) is suitable for the given topology (2).

12. The method (100) according to claim 11, wherein the corresponding input ports (I) of the candidate process modules (24-27) are in particular input ports (I) that provide the same process functionality as the input ports (I) of the at least one process module (21-23), and the corresponding output ports (O) of the candidate process modules (24-27) are in particular output ports (O) that provide the same process functionality as the output ports (O) of the at least one process module (21-23).

13. The method (100) according to claim 12, wherein determining whether a candidate process module (24-27) from the pool (6) of available modules (24-27) is suitable for the given topology (2) further comprises: • determining (124), in a state in which all connections to ports (I, O) of the candidate process module (24-27) have been made according to the given topology (2), whether all ports (I, O) of the candidate process module (24, 27) required for running the utilized services of the candidate process module (24-27) are connected; and • in response to determining (125) that the utilized services of the candidate process module (24-27) cannot be run due to a missing connection to a port (I, O), determining (126) that the candidate process module (24-27) is not suitable for the given topology (2).

14. The method (100) according to any one of claims 1 to 2, further comprising: in response to determining (141) that two or more candidate process modules (24-27) have the same theoretical utilization (24b-27b) that is equal to or higher than the theoretical utilization (21b-23b) of the at least one process module (21-23), selecting (142) the candidate process module (24-27) with the least total number of input ports (I) and output ports (O) as the candidate process module (24-27) to replace the at least one process module (21-23).

15. The method (100) according to claim 1, further comprising: for two or more candidate process modules (24-27) with a theoretical utilization (24b-27b) that is at least as high as the theoretical utilization (21b-23b) of the at least one process module (21-23), evaluating (143) a value of at least one predetermined key performance indicator (7) that would occur in turn in case the at least one process module (21-23) would be replaced by this candidate process module (24-27), and selecting (144) the candidate process module (24-27) with the best result value of the key performance indicator (7) as the candidate process module (24-27) to replace the at least one process module (21-23).

16. The method (100) according to claim 1, wherein: • For a plurality of different process modules (21-23) in the given topology (2), utilization amounts of the at least one resource and / or performance (21a-23a) are specifically obtained (115); • From each of the different process modules (21-23), a respective candidate optimized topology (2') of the modular industrial plant (1) is generated (145); • For each candidate optimized topology (2'), a value of at least one key performance indicator (7) that would occur if the candidate optimized topology (2') were to be implemented is determined (150); and • from the value of the key performance indicator (7) and at least one optimal criterion (8), an optimized topology (2 ) is selected (160) from the candidate optimized topologies (2') 17. The method (100) of claim 15 or 16, wherein the key performance indicator (7) specifically comprises one or more of: a cost of executing the industrial process; a total number of modules in the modular industrial plant (1); a throughput from one or more educts to one or more products of the industrial process as a whole; and an energy consumption of the industrial process.

18. The method (100) of any one of claims 1 to 2, wherein: • the utilization amounts of the at least one resource and / or performance (21a-23a) are specifically obtained (117) for a combination of two or more interconnected process modules (21-23); • the search is specifically performed (127) for candidate process modules (24-27) that can replace the combination of process modules (21-23) and that fit the given topology (2); and • the optimized topology (2 ) of the modular industrial plant is generated (147) by replacing the combination of the two or more process modules (21-23) with a candidate process module (24-27) having the same or higher theoretical utilization (24b-27b) compared to the combination of the two or more process modules (21-23).

19. A computer program comprising machine-readable instructions which, when executed by one or more computers and / or industrial control systems, cause the one or more computers and / or the industrial control systems to perform the method (100) of any one of claims 1 to 18.

20. A non-transitory computer storage medium and / or a download product having the computer program of claim 19.

Images