System for optimising production

Abstract

The invention relates to a system (100) for optimising production, wherein the production is characterised by a production plan, comprising: an ordering means (110), which is designed to output a purchase order, wherein the purchase order characterises an order quantity of resources for the production and a delivery time of an arrival of the ordered order; a planning means (120) which is designed to determine the plurality of planning parameters on the basis of the production plan, wherein the plurality of planning parameters characterises parameters which are designed to implement the production plan; an optimisation means (130), which is designed to create a simulation of the planning means (120) and to optimise a plurality of planning parameters of the simulation by means of a cost function based on input data, wherein the input data comprises at least one planning parameter of the planning means currently used by the planning means (120), wherein the optimisation means (130) is also designed to output a plurality of optimised planning parameters on the basis of the simulation and the cost function, wherein the ordering means (110) is designed to output the purchase order on the basis of the plurality of optimised planning parameters.

Description

[0001] System for optimizing production

[0002] The invention relates to a system for optimizing production and a computer-implemented method for optimizing production.

[0003] The production of a product requires the regular provision of various resources such as materials, components, and the like. These resources must be provided simultaneously or in close succession, depending on the production process, to ensure smooth production. The more complex the product, the more complex the production planning becomes. A wide range of parameters must be considered during planning to enable efficient and cost-effective production.

[0004] At least as many resources as are currently needed for production must be available. Furthermore, safety stocks must be maintained, production plan adjustments factored in, supply bottlenecks considered, and much more. Storing arbitrarily large quantities of resources incurs high storage costs, so an optimal amount of resources should be chosen.

[0005] Production planning is therefore subject to various uncertainties and complex interrelationships. Due to this complexity, it is hardly achievable reliably using familiar, simple methods such as basic material requirements planning. Furthermore, good planning requires knowledge of the current state of the material master data. However, this data is often poorly maintained, which further complicates planning. Inefficient planning and ordering can lead to production outages and / or increased costs, for example, for storage.

[0006] It is an object of the invention to provide a system and a computer-implemented method that improve upon at least one or more of the aforementioned disadvantages. In particular, it is an object of the present invention to determine an optimal production plan. According to a first aspect, this object is achieved by a system for optimizing production, wherein the production is characterized by a production plan. The system comprises an ordering system configured to issue a purchase order, the purchase order specifying an order quantity of resources for production and a delivery date for receipt of the ordered goods. Furthermore, the system comprises a planning system configured to determine the multitude of planning parameters based on the production plan, wherein the multitude of planning parameters are parameters configured to implement the production plan.The system further includes an optimization tool designed to create a simulation of the planning tool and to optimize a variety of planning parameters of the simulation using a cost function based on input data. The input data includes at least one planning parameter currently used by the planning tool. Based on the simulation and the cost function, the optimization tool is further designed to output a variety of optimized planning parameters. The ordering tool is designed to generate the purchase order based on the variety of optimized planning parameters.

[0007] The core of the system combines cost function optimization with simulation to determine a multitude of optimal planning parameters. A planning parameter can be considered optimal if the cost function, using the optimal planning parameter, exhibits the lowest possible or minimum cost value and / or results in costs below a predetermined cost threshold. This allows for the automated determination of one or more optimal purchase orders to ensure production according to the production plan. Based on the optimization tool's planning parameters, the simulation can model current and / or future production behavior. The cost function can be designed such that the associated costs are minimized using the optimization tool.The interplay of simulation and cost function allows for the determination of optimal planning parameters for the production plan. Existing solutions are limited to user input and often user-defined planning parameters. Furthermore, it is virtually impossible for users to simulate such complex production processes and various scenarios with different planning parameters. Consequently, manual effort is significantly reduced. Additionally, the process is considerably simplified for the user. Optimal production is ensured through the optimized planning parameters and the resulting purchase order.

[0008] The ordering tool can be or include an enterprise resource planning (ERP) tool. The ordering tool can be a computer or a computer component and / or a network.

[0009] The planning tool can have one or more planning parameters, which are at least partially maintained and / or predefined manually by a user. Consequently, these planning parameters are limited to the user's experience and judgment.

[0010] The planning tool can be or include a material requirements planning tool. The planning tool can be a computer or a computer component and / or a network component.

[0011] The system's means mentioned can be configured to store computer program segments for carrying out the described steps and for executing the computer program segments.

[0012] The system can be a system connected to and / or linked with a production facility, or a production system for producing based on the production plan.

[0013] The purchase order may contain further information, such as delivery terms, prices, a schedule and / or other information relevant to production.

[0014] The multitude of planning parameters can include one or more of the following: at least one supplier for the delivery of a resource, in particular an address and / or contact address of that supplier; an expected delivery time of the supplier from the time of receipt of the purchase order until the delivery time; current resource master data; a safety stock or safety time of at least one of the resources; a planning horizon of at least one of the resources; a production requirement of the production plan; the supplier's delivery terms; a required quantity of resources based on the production plan; a current and / or future inventory of resources; at least one planned purchase order already issued to the supplier and / or currently being processed by the supplier.The delivery time can characterize the point in time at which the order is available to the ordering party, in this case the system or the system user. Delivery conditions can characterize the framework for a supplier's delivery, for example, a minimum order quantity. The quantity of resources can be the total quantity of resources required according to the production plan to execute the entire production plan. The order quantity can be the specific quantity ordered. The specific quantity ordered can be less than the quantity of resources, for example, because only such a quantity of resources is needed for current production to save on storage costs.

[0015] The input data can include data from other sources. These additional sources can be an SAP system, tracking information from delivery vehicles transporting ordered resources, and / or other data characterizing production, such as Excel files.

[0016] The system can include an initial user interface designed to output at least one piece of information to a user. This information can characterize at least one key performance indicator (KPI) of production based on the multitude of current and / or optimized planning parameters, at least one planning parameter from the multitude of current and / or optimized planning parameters, a comparison between the multitude of current planning parameters and the multitude of optimized planning parameters, and the purchase order based on the multitude of optimized planning parameters. The initial user interface can be an ERP user interface that is connected to and / or linked with the ordering system.

[0017] The first user interface can be configured to issue an optimization start prompt to the user to initiate optimization, to determine optimal purchase order behavior based on the multitude of optimized planning parameters, to request confirmation from the user to confirm at least some of the multitude of optimized planning parameters, and / or to output the multitude of optimized planning parameters to the ordering system, current purchase order behavior, and / or a warning regarding individual resources, suppliers, and / or planning parameters. The optimization start prompt can initiate optimization based on user input. Furthermore, the user can use the optimization start prompt to specify one or more of the optimal planning parameters, so that these are omitted during optimization and considered fixed.The purchase order behavior can be characterized by the issuance of past and / or future purchase orders. A warning can be issued if a resource falls below or exceeds a predetermined minimum and / or maximum quantity, particularly one that is manually set.

[0018] The system may further include a second user interface configured to output at least one of the following to the user: an overview of the temporal progression of a resulting resource inventory based on the multitude of current planning parameters and the multitude of optimized planning parameters; a maximum, minimum, and / or average production range; a range fluctuation; an average resource inventory over a predetermined period; a projected inventory at a predetermined future point in time; or a fluctuation in the future resource inventory. The first and second user interfaces may be interconnected via signaling. Outputs from the first user interface can be sent to the second user interface and vice versa.The second user interface can be configured to display one or more key performance indicators (KPIs) based on the multitude of current planning parameters and / or the multitude of optimized planning parameters. This second user interface can be a management interface and, in particular, be provided to management or associated users.

[0019] The second user interface can be configured to output: an optimization input, which is configured to receive user input to set one or more of the multitude of planning parameters of the simulation, the multitude of optimized planning parameters and / or the cost function of the optimization tool; and / or a policy input, which is configured to receive user input to set supply chain policies to be considered by the optimization tool during optimization.

[0020] The input data can include data from one or the ordering system, in particular the ERP system.

[0021] The optimization tool can be configured to receive one or more predetermined optimized planning parameters from a forecasting tool. The system can include the forecasting tool. The forecasting tool can be configured to predict optimized planning parameters based on past planning parameters, production plans, and / or production data. The received optimized planning parameters can be excluded from the optimization process and / or fixed for the simulation during its creation, so that these optimized planning parameters are not further modified.

[0022] The cost function can characterize cost terms related to resource storage, resource delivery, and / or deviations from target values ​​in the production plan. Further cost terms can include: costs for warehousing; costs for delivery; costs for deviations from a target resource range; costs for falling below a minimum range; costs for deviations from a predetermined service level; costs for deviations from a supply framework agreement for orders; costs for deviations from a previous production plan; and costs for production downtime.

[0023] The cost function can include a weighting of at least one of the cost terms, where the weighting is determined based on fluctuations in resource requirements, delivery time, procurement time, delivery time variation, a predetermined service level, supplier reliability, material category of the resources, and / or uncertainty. Procurement time can characterize the period from the placement of an order until the delivery date.

[0024] To optimize the cost function, an optimization method can be used. Suitable optimization methods include various solution methods based on linear programming, such as mixed-integer linear programming (MILP), or on the basis of constraint optimization, also known as constraint programming (CP), or on the basis of particle swarm optimization, or on the basis of hyperparameter search algorithms. Machine learning methods and heuristics are also suitable.

[0025] The results or the runtime of the optimization can be further improved, among other things, by continuously evaluating the simulation even with a discrete search space and ultimately outputting the next discrete planning parameters. Discrete parameter search spaces can also be interpolated.

[0026] Optimization can be performed automatically. Alternatively or additionally, optimization can be started at regular intervals and / or based on user input. Alternatively or additionally, optimization can be performed if monitoring of the planning and / or ordering resources reveals potential for improvement and / or if a planning parameter deviates from reality.

[0027] Optimization optimizes the cost function, which is composed of several cost terms (at least one). These cost terms can be combined into a single numerical value using a mathematical function and weighted differently. The weighting of the various cost terms can also change over time. The selection of criteria and weighting of the cost function can be done via the second user interface or alternatively, defined by a developer within the optimization process.

[0028] The cost function can be formulated as follows, for example: ...,X n

[0029] The cost function consists of three components, which can be weighted using a, β, and y: inventory costs, costs for deviations from a target range, and costs for falling below a minimum range. The example can be evaluated over a period of, for example, 200 days. The variables of the cost function are defined below:

[0030] • x Q is the inventory on day 0;

[0031] • c0 is the range on day 0;

[0032] • c_target Q is the target range for day 0;

[0033] • c_min is a minimum range;

[0034] • n is the optimization horizon.

[0035] The optimization tool can be configured to create the simulation based on user input, implementing logic and / or an algorithm, and / or at least partially copying logic and / or an algorithm from the planning tool. Consequently, individual or all logics and / or algorithms from the existing planning tool can be copied for the simulation.

[0036] The optimization tool can be further developed to validate and adjust the simulation based on input data, historical system data, and / or synthetic input data being entered into the simulation and the planning tool. To generate the best possible / most accurate simulation, it may be necessary to measure its accuracy and compare it to the original planning tool, for example, using an error or standard deviation. For this comparison, real historical data and / or a specially trained input generator that generates synthetic, valid input data can be used. The advantage of the input generator is its ability to simulate specific edge cases or rare occurrences. Furthermore, the validation or accuracy measurement can be used to blend historical data with data from the input generator.

[0037] The optimization tool can be configured to create a large number of simulations and optimize numerous planning parameters within these simulations using a cost function based on the input data or a cost function assigned to each individual simulation. The number of simulations can vary, at least partially. Furthermore, the optimization tool can select at least one simulation from the large number of simulations to optimize the respective cost function based on validation of that simulation.

[0038] Depending on the specific circumstances, it may be necessary to generate / train multiple simulations instead of just one and store them in a model database. The individual simulations can vary in their complexity and, for example, be adaptively exchanged based on the output of a corresponding quality measurement. Furthermore, learned simulations can be retrained after a certain period if they deviate from the real system, in this case, the planning tool. Another approach is continuous interpolation for a simulation that was trained on discrete input data.

[0039] The system may further include a means of production for manufacturing based on the resources provided by means of the purchase order.

[0040] According to a second aspect, the task is accomplished by a computer-implemented

[0041] A method for optimizing production is solved, where production is characterized by a production plan. A multitude of planning parameters for a purchase order, based on a production plan, are to be determined using a planning tool. These parameters are designed to implement the production plan. The purchase order specifies an order quantity of resources for production and a delivery date for receipt of the ordered goods. The method comprises the following steps:

[0042] - Creating a simulation of the planning tool and optimizing a variety of planning parameters of the simulation using a cost function based on input data, wherein the input data includes at least one planning parameter of the planning tool currently used by the planning tool,

[0043] - Output of a variety of optimized planning parameters based on the simulation and the cost function;

[0044] - Issuing a purchase order based on the multitude of optimized planning parameters.

[0045] Device features described in relation to the system according to the first aspect can be implemented as process features of the process according to the second aspect.

[0046] The task is solved according to a third aspect by a computer program product comprising commands that cause a system according to the first aspect to execute the procedure according to the second aspect.

[0047] Preferred embodiments are explained by way of example with reference to the accompanying figures. These show:

[0048] Fig. 1 shows a schematic representation of a system for optimizing production; and

[0049] Fig. 2 shows a schematic representation of a computer-implemented method for optimizing production. Fig. 1 shows a schematic representation of a system 100 for optimizing the production of, for example, one or more products. Production takes place according to a production plan.

[0050] System 100 includes a purchasing tool 110 configured to issue a purchase order. The purchase order specifies an order quantity of resources for production and a delivery date for receipt of the ordered goods. Purchasing tool 110 can be configured to issue the purchase order to one or more suppliers, as indicated by the arrow pointing left in Fig. 1. The purchase order can specify various resources and / or one or more delivery dates. Multiple purchase orders can be issued using purchasing tool 110, for example, to multiple suppliers and / or a single supplier.

[0051] System 100 also includes a planning tool 120, which is designed to determine a multitude of planning parameters based on the production plan. These numerous planning parameters characterize parameters used to implement the production plan. Such planning parameters can, for example, be derived from a supplier's delivery time or characterize those parameters the supplier requires for delivery after receiving the purchase order. For instance, if a supplier needs 9 days for delivery and the order is required in 14 days, the purchase order can be issued to the supplier in 5 days. These planning parameters of planning tool 120 can be preset by a user. Often, these planning parameters are not updated, which can lead to inaccurate planning.To complicate matters further, production requires a wide variety of resources, such as components, at sometimes different times. Furthermore, a certain safety stock of resources is regularly ordered and / or stored. This safety stock can change with production plan modifications, meaning that some safety stocks may no longer be necessary. However, this setting can be overlooked by the user. Since production planning is generally very complex and difficult, and production is also extremely cost- and resource-sensitive, the inventors of this patent application recognized that System 100, as described below, makes it possible to achieve optimal planning.

[0052] System 100 further comprises an optimization tool 130, which is designed to create a simulation of the planning tool and to optimize a multitude of planning parameters of the simulation using a cost function based on input data. According to Fig. 1, the optimization tool 130 is subdivided into two sections: an optimizer 131 for optimization and the simulation 132. The input data includes at least one planning parameter of the planning tool currently used by the planning tool.

[0053] The simulation can be created manually by the user, for example, by copying logic and / or algorithms from planning tool 120. Optimization tool 130 is further developed based on the simulation and the cost function to output a variety of optimized planning parameters. For this purpose, the cost function, which contains cost terms in conjunction with the planning parameters, is optimized. Optimization can mean, for example, that the costs of individual cost terms or the cost function are below a predetermined cost threshold or are minimal. The input data can include, for example, data from the purchasing tool, material master data, and predetermined optimized planning parameters. The cost function is then optimized based on this input data to determine the optimal planning parameters.The ordering system is further trained to issue the purchase order based on the multitude of optimized planning parameters.

[0054] Figure 1 further shows a first user interface 141. This first user interface 141 can represent an interface between the user, e.g., a planner, and an MRP optimization tool, through which several bidirectional interaction options are available. For example, the user can select resources and materials to be optimized by the optimization tool 130. Furthermore, the currently used planning parameters and the current ordering behavior can be displayed. Once the optimized planning parameters have been determined, the user can transfer the optimal planning parameters via the first user interface 141 so that they are output to the ordering tool 110.

[0055] Figure 1 shows a second user interface 142. This second user interface 142 can represent the interface between higher-level users (e.g., logistics managers) and MRP optimization, through which several bidirectional interaction possibilities are mapped. For example, key performance indicators (KPIs) can be displayed, which are achieved based on the currently used and / or optimized planning parameters. Furthermore, a supply chain policy can be set up, which the optimization tool 130 must consider in the cost function.

[0056] Consequently, the proposed System 100 enables optimized planning and thus also optimized production. The user is also supported in achieving optimal production. Additionally, outdated settings for planning parameters can be updated. System 100 can be a production system. System 100 can be a production resource for manufacturing based on the resources provided by the purchase order.

[0057] Fig. 2 shows a computer-implemented method 200 for optimizing production, which can be executed by the system 100. The system 100 can include a memory in which the method 200 is stored in the form of a computer program.

[0058] A multitude of planning parameters for a purchase order are to be determined based on a production plan using a planning tool 110. These numerous planning parameters are defined as those used to implement the production plan. The purchase order specifies a quantity of resources for production and a delivery date for the receipt of the ordered goods. The procedure 200 comprises creating a simulation of the planning tool 120 and optimizing a multitude of planning parameters from the simulation using a cost function based on input data. The input data includes at least one planning parameter of the planning tool 120 that is currently used by the planning tool.The procedure 200 further includes an output 220 of a large number of optimized planning parameters based on the simulation and the cost function and an output 230 of a purchase order based on the large number of optimized planning parameters.

[0059] Using the proposed method 200, optimal production can be achieved identically to system 100.

[0060] Reference mark

[0061] System for optimizing production

[0062] Ordering materials

[0063] Planning resources

[0064] Optimization tools

[0065] optimizer

[0066] Simulation first user interface second user interface computer-implemented method for optimizing production Creating a simulation of the planning tool and optimizing a variety of planning parameters of the simulation

[0067] Output of a large number of optimized planning parameters based on the simulation and the cost function.

[0068] Issuing a purchase order based on the multitude of optimized planning parameters.

Claims

Patent claims 1. System (100) for optimizing production, wherein the production is characterized by a production plan, comprising: an ordering device (110) configured to issue a purchase order, wherein the purchase order characterizes an order quantity of resources for production and a delivery time of receipt of the ordered goods; a planning device (120) configured to determine the multitude of planning parameters based on the production plan, wherein the multitude of planning parameters characterize parameters configured to implement the production plan;An optimization tool (130) configured to create a simulation of the planning tool (120) and to optimize a variety of planning parameters of the simulation using a cost function based on input data, wherein the input data includes at least one planning parameter of the planning tool (120) currently used by the planning tool (120), wherein the optimization tool (130) is further configured, based on the simulation and the cost function, to output a variety of optimized planning parameters, and wherein the ordering tool (110) is configured to output the purchase order based on the variety of optimized planning parameters.

2. System (100) according to claim 1, wherein the plurality of planning parameters comprises one or more of the following planning parameters: at least one supplier for the supply of a resource; an expected delivery time of the supplier from the time of receipt of the purchase order until the delivery time; current resource master data; a safety stock or safety time of at least one resource; a fixing horizon of at least one resource; a production requirement of the production plan; The supplier's delivery terms; a required quantity of resources based on the production plan; a current and / or future inventory of resources; at least one planned purchase order already issued to the supplier and / or currently being processed by the supplier.

3. System (100) according to claim 1 or 2, further comprising: a first user interface (141) configured to output at least one piece of information to a user, wherein the information characterizes at least one key performance indicator of production based on the multitude of current and / or optimized planning parameters, at least one planning parameter of the multitude of current and / or optimized planning parameters, a comparison between the multitude of current planning parameters and the multitude of optimized planning parameters, and the purchase order based on the multitude of optimized planning parameters.

4. System (100) according to claim 3, wherein the first user interface (141) is configured to issue an optimization start request to the user to start the optimization, an optimal order order behavior based on the plurality of optimized planning parameters, a confirmation request to the user to confirm at least part of the plurality of optimized planning parameters and / or to output the plurality of optimized planning parameters to the ordering means (110), an current order order behavior and / or a warning about individual resources, suppliers and / or planning parameters.

5. System (100) according to one of the preceding claims, further comprising: a second user interface (142) configured to output at least one of the following to the user: an overview of a temporal progression of a resulting inventory of resources based on the multitude of current planning parameters and the multitude of optimized planning parameters; a maximum, minimum and / or average production range; a fluctuation in range; an average resource inventory over a predetermined period; a projected inventory at a predetermined future point in time; a fluctuation in the future resource inventory.

6. System (100) according to claim 5, wherein the second user interface (142) is configured to output: an optimization input configured to receive user inputs from the user to set one or more of the plurality of planning parameters of the simulation, the plurality of optimized planning parameters and / or the cost function of the optimization tool (130); and / or a policy input configured to receive user inputs from the user to set supply chain policies to be taken into account by the optimization tool during optimization.

7. System (100) according to one of the preceding claims, wherein the optimization means (130) is configured to receive one or more predetermined optimized planning parameters from a forecasting means, wherein the received optimized planning parameters are excluded from optimization and / or are specified for the simulation when creating the simulation.

8. System (100) according to one of the preceding claims, wherein the cost function characterizes cost terms with respect to the storage of resources, the delivery of resources and / or a deviation from target values ​​to be achieved in the production plan.

9. System (100) according to one of the preceding claims, wherein the cost function has a weighting of at least one of the cost terms, wherein the weighting is determined based on a fluctuation in resource requirements, delivery time, procurement time, fluctuation in delivery time, a predetermined service level, supplier reliability, material category of resources and / or uncertainty, wherein the procurement time characterizes a period from the issuance of the order to the delivery time.

10. System (100) according to one of the preceding claims, wherein the optimization means (130) is configured to create the simulation based on a user input from the user to implement a logic and / or an algorithm and / or to at least partially copy a logic and / or an algorithm of the planning means.

11. System according to one of the preceding claims, wherein the optimization means is further configured to validate and adapt the simulation based on inputting input data, historical data of the system and / or synthetic input data into the simulation and into the planning means (120).

12. System (100) according to one of the preceding claims, wherein the optimization means (130) is configured for creating a plurality of simulations and for optimizing plurality of planning parameters of the plurality of simulations by means of a cost function based on the input data or a cost function assigned to the respective simulation, wherein the plurality of simulations are at least partially different, wherein the optimization means (130) is further configured for selecting at least one simulation of the plurality of simulations for optimizing the respective cost function based on a validation of the respective simulation.

13. System (100) according to any of the preceding claims, further comprising: a production means for producing based on the resources provided by means of the purchase order.

14. Computer-implemented method (200) for optimizing production, wherein the production is characterized by a production plan, wherein a variety of planning parameters for a purchase order are to be determined based on a production plan for production by means of a planning tool (120), wherein the variety of planning parameters are parameters designed to implement the production plan, wherein the purchase order characterizes an order quantity of resources for production and a delivery time of receipt of the ordered order, the method comprising the steps: Creating (210) a simulation of the planning tool (120) and optimizing a variety of planning parameters of the simulation using a cost function based on input data, wherein the input data includes at least one planning parameter of the planning tool (120) currently used by the planning tool (120), Output (220) a variety of optimized planning parameters based on the simulation and the cost function; Output (230) of a purchase order based on the multitude of optimized planning parameters.

15. Computer program product comprising instructions that cause a system (100) according to any one of claims 1 to 13 to execute the method (200) according to claim 14.

Images