Method and system for changeover analysis for improving overall equipment effectiveness

Abstract

A method for analyzing and optimizing changeover events in production processes of a manufacturing system by a smart algorithms (compute which uses data captured by a standalone Overall Equipment Effectiveness (OEE) measurement system, the following steps of Identifying minimum data requirements and extracting them from the captured data, in particular a manufacturing equipment name, timestamps for the start and end of a task, a description of an executed task, at least one manufacturing batch number, a quantity number for produced products in a production event and available operator remarks; Sorting the extracted minimum data requirements for the identified changeover events by the computer through using a lookup- table which contains standard acronyms which are defined by operations for a changeover class or availability loss categories, resulting in a data set for each category; Analyzing every data set with regard to the identified changeover events for different scenarios with different production process conditions, wherein the data in all datasets is assembled in chronological order for an equipment; Optimizing the data by the computer through assigning a changeover category to the identified changeover events via a look-up table containing a list of key changeover types, merging any identified sub-sequent changeover events to a final changeover event, and assigning all available associated user-defined attributes to the identified and analyzed changeover events; Visualizing the data to a user via a user interface; and Adapting the production processes in the manufacturing system to improve its operational efficiency by the user, comprising.

Description

[0001] Method and System for Changeover Analysis for improving Overall Equipment Effectiveness

[0002] The hereby described invention discloses a method and system for analyzing changeovers in a manufacturing process to perform Overall Equipment Effectiveness (OEE) calculations to improve said manufacturing process.

[0003] Technical Field

[0004] This invention deals with the digital shopfloor management in any manufacturing process running in a batch production like biopharmaceutical and specialty chemicals or continuous production like automotive assembly lines.

[0005] Background and description of the prior art

[0006] The present invention relates to a method and system for analyzing changeovers in a manufacturing process to support Overall Equipment Effectiveness (OEE) calculations. OEE measurement systems are widely used in the manufacturing industry to monitor equipment data, translate it to key performance indicators (KPIs), and optimize processes for maximum throughput and capacity utilization. However, the current approach does not provide a systematic analysis of changeover events, which can account for up to a half of equipment capacity loss in batch manufacturing processes depending on complexity of the manufacturing process, equipment, and level of acceptance.

[0007] Changeover is a complex event that involves a series of sub-tasks, with the duration varying depending on the complexity and compliance requirements. In some cases, changeovers can take several hours, resulting in a significant loss of production time. The current OEE measurement systems capture the task details of a changeover event but are unable to distinguish between technical breakdowns, trial runs, start-up losses, and other factors that can affect equipment availability. This results in a lack of transparency and makes it difficult to analyze the overall effective utilization of manufacturing equipment.

[0008] The challenge of analyzing vast amounts of manufacturing data, which generate millions of data points daily, is a significant obstacle. The first hurdle is identifying when a changeover event is taking place. Once identified, it is necessary to sort and remove the noise from the changeover data, precisely capture relevant sub-tasks, calculate total time spent during the event, and overcome human errors in manual data entry. Additionally, the correct type / category of changeover must be assigned, and any operator comments should be made available for deep dives into the data later.

[0009] The lack of a standardized solution has led to an approach that varies from user to user and is dependent on individual operational experience. As a result, it is challenging to integrate the data within operational Key Performance Indicators (KPIs) and is considered less reliable than a standardized solution. Manual analysis with semi-efficient tools takes hours daily, making it difficult to leverage changeover data for benchmarking and leaving a considerable portion of manufacturing optimization potential unattended.

[0010] Optimizing the changeover phases is one of the major stages that organizations would like to improve. Efficient data identification, extraction, processing, and representation are needed to represent inefficiencies in an easy-to-understand manner, allowing organizations to take the necessary steps to optimize equipment efficiencies (OEE) and covert significant portion of non-value adding activities into value adding production time. Without measurement, improvement is impossible, which is why users need to be equipped with data for statistical process controls, trend analysis, and deep dives into operational challenges, adhering to standard takt times and applying lean and six sigma tools to further improve the process.

[0011] There is a clear need for a standardized approach to analyze changeover information, generate insights, and equip users with data analytics for statistical analysis to maximize overall equipment effectiveness.

[0012] Summary of the invention

[0013] The task of this patent application is therefore to find an improved method to analyze, monitor and improve the operational performances leveraging data- driven decisions which can overcome the known limitations of the prior art.

[0014] This task has been solved by a method for analyzing and optimizing changeover events in production processes of a manufacturing system by a computer which uses data captured by a standalone Overall Equipment Effectiveness (OEE) measurement system, the following steps comprising: Identifying minimum data requirements and extracting them from the captured data, in particular a manufacturing equipment name, timestamps for the start and end of a task, a description of an executed task, at least one manufacturing batch number, a quantity number for produced products in a production event and available operator remarks, and using the extracted minimum data requirements to identify changeover events; Sorting the extracted minimum data requirements for the identified changeover events by the computer through using a look-up table (which could be for instance be provided by operational experts) which contains standard acronyms which are defined by operations for a changeover class or availability loss categories, resulting in a data set for each category; Analyzing every data set with regard to the identified changeover events for different scenarios with different production process conditions, wherein the data in all datasets is assembled in chronological order for an equipment; Optimizing the data by the computer through assigning a changeover category to the identified changeover events via a look-up table containing a list of key changeover types; Merging any identified sub-sequent changeover events to a final changeover event, and assigning all available associated user-defined attributes to the identified and analyzed changeover events; and Visualizing the data to a user via a user interface; and adapting the production processes in the manufacturing system to improve its operational efficiency, comprising. This proposed method overcomes the mentioned challenges by using advanced algorithms to identify changeover events, remove noise and other availability losses, assign the right type / category of changeover, and consolidate information for further analysis. By doing so, the method enable users to identify improvement opportunities and convert non-value adding time to value-adding production time, thereby increasing overall equipment throughput and capacity utilization.

[0015] The solution developed is analyzing the data captured in any standalone OEE measurement system. The solution has five phases to deliver the expected outcome visualizing the data with user-friendly dashboards.

[0016] In a first phase the minimum data requirements are identified. The source of raw data is any standalone OEE measurement system installed on any equipment. This system is capturing the basic information about the production status of an equipment which is later assigned to specific availability loss categories by the equipment operators like, technical breakdown, organizational stoppages, changeovers and wherever possible sub-categories and remarks to justify deviations if any. To execute changeover analysis, the solution needs the mentioned minimum requirements, for example:

[0017] Manufacturing equipment name

[0018] Timestamps with start and end of a task Description of a task executed Manufacturing batch number

[0019] Quantity produced in a production event

[0020] Operator remarks if any

[0021] The data to be analysed is ingested into the analytics algorithms at any interval defined by the end-users, continuous, hourly, daily, weekly, etc.

[0022] In Phase two the manufacturing data is sorted. Whenever there is a change in manufacturing batch number, a changeover is executed. When a batch number is changing, time interval between last good part produced for last batch / production order in any previous production event to next good part produced with next batch / production order number is typically a changeover event. To ease the analysis, a lookup table containing standard acronyms defined by operations for changeover class or higher availability loss category is maintained. In some cases where a batch is broken down with an intermittent changeover to meet the compliance or operational requirements, a change on batch number is not seen. Both type of data is analyzed separately in multiple loops in following phase. To ease the analysis, such events are identified, and only relevant data is considered for further analysis while rest of the data is discarded. While doing so chronological order of data events is maintained.

[0023] Phase three results in the changeover data analysis itself and is the core of the solution and executed in one respective scenario for each of the data set categories from phase 2. It with fine-tuning the outcome accommodating the ramp-up (start-up) challenges for the next batch. Each scenario data is analyzed separately for each data set category and later all are combined chronologically:

[0024] Scenario 1 : Changeover sub-tasks without any interruptions. Here all changeover activities are running smoothly without any technical or organizational breakdowns or any ramp-up challenges. In such a case, difference between end of last production event of previous batch to start of first production event of next batch is considered as changeover duration.

[0025] Scenario 2: Changeover with noise of other breakdowns or stoppages.

[0026] Like scenario 1 , changeover duration is calculated but here noise duration from breakdowns, stoppages, organizational or planned, is discounted.

[0027] Scenario 3: Data similar to scenario 2 but with few production events less than a defined time interval. This scenario represents the typical challenges when a center-lining activity is not executed in defined standards and equipment is facing ramp-up challenges. During such a ramp-up of an equipment following a changeover over, one could see several minor events where the equipment status frequently changes to production but actual production event producing acceptable good parts has not yet started. Here one need to have a solid historical data at hand to define the production time threshold which if passed, equipment is considered to be running smoothly. In the first pilot, for example we executed for 20 manufacturing equipment’s with data of 6 months and over 10,000 changeover events, a threshold of 100 seconds determined meeting 3-sigma confidence limit of 99.7%. This value is a user defined precision and should be updated at a defined time interval matching data quality standards. The time interval between last production event of the previous batch till this threshold value is overtaken, excluding the noise is actual changeover duration. On demand, data from scenario 3 could be further investigated to optimize on ramp-up challenges.

[0028] At the end of this phase, algorithms generate 7 data sets; 3 scenarios for each of 2 data sets considering batch number changing or not and any remaining data which is considered as human error. The human-error data set is reported back to the key user to improve data quality and advised to have almost no data in this data set. In Phase 4 the data is further optimized to meet expected quality meeting end-user requirements for visualization. This phase is about putting data from all 6 useful datasets together in chronological order for an equipment. Data optimization is executed in 3 steps:

[0029] Step 1 : Assign right category to changeover event.

[0030] It is quite often possible that in a changeover event, multiple changeover types are assigned to different timestamps. Challenge here is to identify the right type and allocate it that specific event. In many manufacturing operations, 4-eye principle is applied where before giving a green signal to produce next batch, an expert or skilled external operator is checking if the changeover is executed properly, if right type is assigned in the system and any compliance procedures are followed as instructed. Considering the manufacturing network in focus executing over 100,000 changeovers where this invention is invented and applied first, the 4-eye principle is actively applied and always practiced before starting a new manufacturing batch. Cases with more than one changeover type assigned are not common but still possible to happen in up to 5% of the events, especially when a complex changeover is executed or a changeover running multiple production shifts or when equipment faces severe breakdown or changes in priority and left unattended for a longer time. In such cases, latest changeover type is assigned to the event. To feed the algorithms a look-up table containing list of key changeover types is maintained with up-to-date operational information.

[0031] Step 2: Combine any sub-sequent changeover events.

[0032] After executing step 1 of phase 4, all data sets that are organized chronologically and have one entry for each changeover event are analyzed if end timestamp of an event is matching with start timestamp of next event. In such a case, such events are merged assigning the name of latest changeover type to the final event. This will ensure to consider any changeovers executed but at last moment manufacturing priorities have changed and equipment went through additional changeovers or in rare cases, changeovers executed during a longer planned stoppages.

[0033] Step 3: Assign additional attributes-shift, product name, remarks.

[0034] Here all associated user-defined attributes which are helpful for deep dives are assigned to analyzed changeovers. All free text remarks entered by operators during a complete changeover event are combined and assigned the final output of the analysis.

[0035] In Phase 5 the data is then visualized. Once all above phases are successfully executed, the output data is ready for visualization and deep dives with any user interface. Users could use any tool for customized visualization, for example, Power Bl, Aera, tableau, MS Excel, Minitab, etc. to name a few. The visualization interface should meet the requirement to drive an effective performance dialogue with possibility to deep dive problem solving.

[0036] The solution to the operational problem improving changeover performance transparency is to identify the various data patterns to deliver simple, easy, understandable datasets that could be represented on a dashboard for business users to take appropriate decisions to improve savings and improve their operational efficiencies.

[0037] Overall, this invention provides a novel and practical solution to analyze changeover events systematically, which is useful for any organization seeking to optimize their manufacturing processes and improve OEE performances supporting operational productivity. This method and system enable manufacturing organizations to analyze changeover events systematically, identify improvement opportunities, and optimize their processes. The proposed method and system are a practical solution that can be easily integrated into the existing IT infrastructure of any manufacturing organization.

[0038] The proposed method and system represent a significant innovation in the field of manufacturing and will have a significant impact on the overall equipment effectiveness of any organization that implements it.

[0039] Advantageous and therefore preferred further developments of this invention emerge from the associated subclaims and from the description and the associated drawings.

[0040] One of those preferred further developments of the disclosed method comprise that the system captures basic information about the production status of an equipment that is later assigned to specific availability loss categories. This allows for a more detailed and accurate analysis of the manufacturing process.

[0041] Another one of those preferred further developments of the disclosed method comprise that the assigned specific availability loss categories could be technical breakdown, organizational stoppages, changeovers, and further sub-categories of deviations. These are only a few examples of the most important types. The invention is not limited to those examples.

[0042] Another one of those preferred further developments of the disclosed method comprise that the changeover event is identified by a change in the manufacturing batch number from the extracted minimum data requirements. The identification is done by the computer using smart algorithms through analyzing the already identified minimum requirement data from the manufacturing data captured by the OEE system. If the batch number in a new data set is has changed, a changeover event has happened. Another one of those preferred further developments of the disclosed method comprise that any remaining data not assigned to a dataset is considered as human-error data and put into a human-error dataset which is reported back to the user to enable the user to improve the captured data quality. It is advised for the user to have almost no data in this human-error data set, by ensuring to improve the input manufacturing data from the OEE system if any human-error dataset is reported back. This level of data transparency helps to close the feedback loop on data quality which is crucial for fact-based decisions and thus, moving the operations in right direction rather than going with wild guess on performances.

[0043] Another one of those preferred further developments of the disclosed method comprise that the merging of the chronologically organized subsequent changeover events to a final changeover event is done by checking via the computer and algorithms if the end timestamp of an event is matching with start timestamp of the next event and in such a case, the events are merged together assigning the name of latest changeover type to the final event and assigning additional attributes such as shift, product name, and remarks as well. This will ensure to consider any changeovers executed but at last moment manufacturing priorities have changed and equipment went through additional changeovers or in rare cases, changeovers executed during a longer planned stoppages.

[0044] Another one of those preferred further developments of the disclosed method comprise that the user interface comprises of a dashboard displayed on a screen connected to the computer. This feature allows for easier understanding and interpretation of the data, as well as more effective decision-making.

[0045] Another one of those preferred further developments of the disclosed method comprise that the data to be analyzed is ingested into the analytics algorithms at any interval defined by the end-users, such as continuous, hourly, daily, or weekly. This allows that the to be analyzed data can be ingested into the analytics algorithms at any interval defined by the user.

[0046] Another one of those preferred further developments of the disclosed method comprise that the changeover events are analyzed based on fine- tuning the outcome accommodating the ramp-up challenges for the next batch. This scenario represents the typical challenges when a center-lining activity is not executed in defined standards and equipment is facing ramp- up challenges. During such a ramp-up of an equipment following a changeover over, one could see several minor events where the equipment status frequently changes to production but actual production producing acceptable good parts has not yet started. This is helpful to generate transparency on technical reliability of manufacturing equipment, effective settings on equipment and first-time right maturity in operations.

[0047] Another one of those preferred further developments of the disclosed method comprise that the changeover sub-tasks without any interruptions are analyzed based on the difference between the end of the last production event of the previous batch to start the first production event of the next batch as their changeover duration. Here all changeover activities are running smoothly without any technical or organizational breakdowns or any ramp-up challenges. In such a case, difference between end of last production event of previous batch to start of first production event of next batch is considered as changeover duration.

[0048] Another one of those preferred further developments of the disclosed method comprise that changeover scenarios with noise of other breakdowns or stoppages are discounted from their changeover duration. This scenario is similar to the situation where all changeover activities are running smoothly without any technical or organizational breakdowns or any ramp-up challenges, but here the changeover duration is calculated but here noise duration from breakdowns, stoppages (organizational or planned) is discounted. This data quality improvement approach integrated within the algorithms helps to understand non-value-added activities performed during a changeover event and if well managed within operational optimization processes, could reveal another opportunity taking operations to next level realizing top industry benchmark performances.

[0049] Another one of those preferred further developments of the disclosed method comprise that the data similar to the changeover scenarios with noise and a few production events less than a defined time interval are analyzed based on a threshold value being overtaken to calculate the actual changeover duration. This scenario often represents the typical challenges that can occur when a center-lining activity is not executed in defined standards and equipment is facing ramp-up challenges. This six- sigma approach to continuously upgrade the algorithms with realistic operational performances is key step in the direction of self-optimizing manufacturing processes and could be a foundation towards developing digital shopfloor management towards Industry 4.0 standard practices.

[0050] Another one of those preferred further developments of the disclosed method comprise that data sets are generated separately for each changeover scenario. This allows for better organization and easier analysis of the data, according to different changeover scenarios.

[0051] Another one of those preferred further developments of the disclosed method comprise that the user-defined attributes assigned in the optimization phase are used for deep dives and further problem-solving by the user. This is helpful in providing additional context and insights that allow for more in-depth analysis of the data which could be further leveraged for data-driven decisions while intra-network product portfolio transfers. Another one of those preferred further developments of the disclosed method comprise that the computer learns from analyzed data through algorithms or human intervention to identify production ramp-up thresholds with a six sigma approach and generate data on understanding the production ramp-up challenges, which are then used to address operational capability and existing competency gaps.

[0052] A further solution to the provided task is a standalone OEE measurement system comprising of the following components of all existing hardware production equipment, a data-based machine learning model component performed by a software, a database, at least one computer to operate the software, wherein all system components are connected a network structure with the computer and the system is configured to capture basic information about the production status of an equipment that is later assigned to specific availability loss categories and used in the method of analyzing changeover events in manufacturing systems described in any of the previous claims. As described already for the underlying method approach, this allows for a more detailed and accurate analysis of the manufacturing process, leading to improved efficiency and productivity.

[0053] A further solution to the provided task is a non-transitory computer-readable medium containing instructions for executing the method of analyzing changeover events in manufacturing systems described in any of the previous claims. This provides flexibility on storing the instructions for executing the method and allows for more efficient implementation of the method.

[0054] The invention with its presented embodiment provides several advantages compared to the known prior art. Those include:

[0055] A faster model building to achieve quickly a first robust monitoring with less time needed to build a model. • The data model is not dependent from a given process, media, scale or manufacturing or any cell line.

[0056] • A better correlation and robustness of the model for a given component.

[0057] • A reduction of operation time, consumable, resources, equipment immobilization.

[0058] Detailed description of the invention

[0059] The method and the system according to the invention and functionally advantageous developments of those are described in more detail below with reference to the associated drawings using at least one preferred exemplary embodiment. In the drawings, elements that correspond to one another are provided with the same reference numerals.

[0060] The drawings show:

[0061] Figure 1 : an example architecture for a non-Hadoop computing environment where the invention logic and algorithms are deployed

[0062] Figure 2: the used algorithm scenarios in a simplified version

[0063] Figure 3: a general schematic about the five phases of the invented method

[0064] Figure 4: a step-by-step illustration of a preferred working example of the invented method in daily operations showing the OEE output and resulting functional loss trends

[0065] Figure 5a: a first part of the step-by-step illustration showing the more detailed changeover analyzation data allowing further deep dives Figure 5b: a second part of the step-by-step illustration showing the more detailed changeover analyzation data allowing further deep dives

[0066] Figure 6: an illustrative example of processing operational data of a changeover event with changeover analytics to deliver simplified output which could be later used by visualization and deep-dive tools presented with figure 4, 5a and 5b.

[0067] Currently this working example is deployed, production - live and working efficiently in an internal datacenter on a Hadoop cluster and is having the tech-stack as: Cloudera-Hadoop 3.1 .x, HDFS, Hive 3.1.x, Apache-Spark 2.x, oozie, etc. The data is processed on every weekend and provided to an Aera platform (visualization user interface) for dashboarding purposes. It can be also realized by using a non-Hadoop computing environment in a further preferred embodiment where the invention logic and algorithms, realized on form of a data model, could be deployed and similar outcome could be obtained, as it is shown in Figure 1 . The hardware for the disclosed system of Figure 1 comprises of a standard or industrial personal computer or server or any other suitable device, especially if the local control unit provides the data model itself, because then a higher processing power as usually provided by a microcontroller is required. In another preferred embodiment the data model is provided by a suitable separate computer with the same or another software at a remote location via a data network using a cloud-based service or a USB flash drive.

[0068] The disclosed invention for changeover analytics is in particular supporting operations to:

[0069] • Improve transparency on impact of internal and external factors on the function losses

[0070] • Opportunity identification and scoping on impactful C / O types or challenging Stock Keeping Units (SKUs) • Track the improvement trends of Single-Minute Exchange of Dies (SMED), planning and operational optimizations

[0071] Figures 4 and 5a, 5b are providing a step-by-step illustration of a preferred working example for an application of the disclosed invention in daily operations. Figure 4 shows thereby the data capturing from the OEE and its subsequent functional loss trends.

[0072] With this high-level overview from Figure 4, the users can continue to further deep dive and investigate the trends, by using the output from the system showing performance deviations by products, at shift level, narrow down and scope for impactful opportunities. Figure 5a and 5b show an example of such an analysis result, provided by the Data Model, its supporting software and computer and shown to the user via the Graphical User Interface in form of a Dashboard displayed on a suitable screen, which is connected to the computer.

[0073] Figure 6 illustrates in a simplified manner, processing of operational raw data collected with any OEE measurement tool to deliver qualitative outputs leveraging smart changeover analytics algorithms developed within this invention.

Claims

Patent claims1. A method for analyzing and optimizing changeover events in production processes of a manufacturing system by a computer which uses data captured by a standalone Overall Equipment Effectiveness (OEE) measurement system, the following steps comprising:• Identifying minimum data requirements and extracting them from the captured data, in particular a manufacturing equipment name, timestamps for the start and end of a task, a description of an executed task, at least one manufacturing batch number, a quantity number for produced products in a production event and available operator remarks, and using the extracted minimum data requirements to identify changeover events;• Sorting the extracted minimum data requirements for the identified changeover events by the computer through using a lookup-table which contains standard acronyms which are defined by operations for a changeover class or availability loss categories, resulting in a data set for each category;• Analyzing every data set with regard to the identified changeover events for different scenarios with different production process conditions, wherein the data in all datasets is assembled in chronological order for an equipment;• Optimizing the data by the computer through assigning a changeover category to the identified changeover events via a lookup table containing a list of key changeover types, merging any identified sub-sequent changeover events to a final changeover event, and assigning all available associated user-defined attributes to the identified and analyzed changeover events;• Visualizing the data to a user via a user interface; andAdapting the production processes in the manufacturing system to improve its operational efficiency by the user.

2. The method of claim 1 , wherein the system captures basic information about the production status of an equipment that is later assigned to specific availability loss categories.

3. The method according to claim 2, wherein the assigned specific availability loss categories are technical breakdown, organizational stoppages, changeovers, and further sub-categories of deviations which are standard sub-elements of an OEE measurement system.

4. The method according to anyone of the previous claims, wherein the changeover event is identified by a change in the manufacturing batch number from the extracted minimum data requirements.

5. The method according to anyone of the previous claims, wherein any remaining data not assigned to a dataset is considered as human-error data and put into a human-error dataset which is reported back to the user to enable the user to improve the captured data quality.

6. The method according to anyone of the previous claims, wherein the merging of the chronologically organized sub-sequent changeover events to a final changeover event is done by checking via the computer using algorithms if the end timestamp of an event is matching with start timestamp of the next event and in such a case, the events are merged together assigning the name of latest changeover type to the final event and assigning additional attributes such as shift, product name, and remarks as well.

7. The method according to anyone of the previous claims, wherein the user interface comprises of a dashboard displayed on a screen connected to thecomputer or any shopfloor management tool leveraged for a user interface and interpretation of the performance data.

8. The method according to anyone of the previous claims, wherein the data to be analyzed is ingested into the analytics algorithms at any interval defined by the end-users, such as continuous, hourly, daily, or weekly.

9. The method according to anyone of the previous claims, wherein the changeover events are analyzed based on fine-tuning the outcome accommodating the ramp-up challenges for the next batch.

10. The method according to anyone of the previous claims, wherein the changeover sub-tasks without any interruptions are analyzed based on the difference between the end of the last production event of the previous batch to start the first production event of the next batch as their changeover duration.11 . The method according to claim 10, wherein changeover scenarios with noise of other breakdowns or stoppages are identified and are discounted from their changeover duration.

12. The method according to claims 10 or 11 , wherein the data similar to the changeover scenarios with noise and a few production events less than a defined time interval are analyzed based on a threshold value being overtaken to calculate the actual changeover duration.

13. The method according to anyone of the previous claims, wherein data sets are generated separately for each changeover scenario.

14. The method according to anyone of the previous claims, wherein the user- defined attributes assigned in the optimization phase are used for deep dives and further problem-solving by the user.

15. The method according to anyone of the previous claims, wherein the computer learns from analyzed data through algorithms or human intervention to identify production ramp-up thresholds with a six sigma approach and generate data on understanding the production ramp-up challenges, which are then used to address operational capability and competency gaps.

16. A computer-readable storage medium containing instructions for executing the method of analyzing changeover events in manufacturing systems described in any of the previous claims.

17. A standalone OEE measurement system configured to capture basic information about the production status of an equipment that is later assigned to specific availability loss categories and used in the method of analyzing changeover events in manufacturing systems described in any of the previous claims.

Images