One approach to addressing these challenges is to use plant or manufacturing intelligence systems.
Given the problematized communication between manufacturing execution systems (MES), plant automation, and enterprise applications, a new breed of applications are coming from the likes of former Lighthammer (now part of SAP), Kinaxis (formerly Webplan), Activplant, and Informance. They offer middleware analytical applications called manufacturing intelligence or plant intelligence that target other applications used to generate corporate-wide visibility of key performance indicators (KPI). These plant portal applications consolidate data taken from a wide range of computing sources—from plant floors, enterprise systems, databases, and elsewhere—and organize these data into meaningful, roles-based information, aggregating the data from disparate sources for analysis and reporting. Connections can through extensible markup language (XML), or open database connectivity (ODBC) standards, with communications managed by a protocol layer in the portal's Web server architecture.
Near real time visibility and transactional exchanges have to be created between enterprise applications and the plant floor with appropriate drill-downs to contextualize and understand the impact of specific manufacturing events. These products are applied to critical plant processes, and monitor production and provide the input required for calculating key metrics, such as overall equipment effectiveness (OEE). In order to increase OEE, data generated by equipment in a production line is acquired and aggregated (preferably in automatically, see The Why of Data Collection).
Information is contextualized using business rules and user roles to create and maintain consistent functional and operational relationships between data elements from these disparate sources. For example, these products can demonstrate the relationship between allowed process variables and ranges of time series-based quality and yield data. It can also analyze information by using business rules to transform raw process data into meaningful KPIs. Data can also be filtered for any noise/outliers; visualized with a context-based navigation and drill-down capabilities; and presented or propagated to determine the factors and root cause disturbances that slow production or impact quality. Ultimately plant-level systems allow decisions to be made that will speed up throughput and increase first-run production.
How It Works
Configuring data sources for integration can be done through templates, which is analogous to selecting a printer for an office application. The real trick, however, is in having sound plant-level models, which are frameworks that portray accurate plant-level context and data management, within the application sets. In manufacturing, even small changes to a master plan can create a so-called "reality gap" and these are historically addressed by last-minute panicking and scrambling, all the while, the business protagonists are not always (if ever) conscious of the impact or even the validity of their "educated guess" decisions.
Thus, these new software applications make it possible to model the cascading consequences of anything users do in response to an unplanned event (like a customer doubling an order or a machine breaking down), which in turn, makes it possible to understand how the other, intertwined parts of the user organization and supply chain will be impacted by a change (see Bridging the Reality Gap Between Planning and Execution).
When users have information about unplanned events and how their responses will impact the company, they should have manufacturing intelligence that can guide them through the forking paths of exception-based, decision-making. The value of the plant-level information indeed changes when enterprises use it to support higher-level, strategic, and tactical business processes. For example, data generated for a department supervisor or for management purposes has one value, and the same data used for Sarbanes Oxley Act (SOX) compliance has another (see Attributes of Sarbanes-Oxley Tool Sets). Moreover, the value of quality assurance (QA) information increases substantially when used to support enterprise-wide warranty issues.
Given the problematized communication between manufacturing execution systems (MES), plant automation, and enterprise applications, a new breed of applications are coming from the likes of former Lighthammer (now part of SAP), Kinaxis (formerly Webplan), Activplant, and Informance. They offer middleware analytical applications called manufacturing intelligence or plant intelligence that target other applications used to generate corporate-wide visibility of key performance indicators (KPI). These plant portal applications consolidate data taken from a wide range of computing sources—from plant floors, enterprise systems, databases, and elsewhere—and organize these data into meaningful, roles-based information, aggregating the data from disparate sources for analysis and reporting. Connections can through extensible markup language (XML), or open database connectivity (ODBC) standards, with communications managed by a protocol layer in the portal's Web server architecture.
Near real time visibility and transactional exchanges have to be created between enterprise applications and the plant floor with appropriate drill-downs to contextualize and understand the impact of specific manufacturing events. These products are applied to critical plant processes, and monitor production and provide the input required for calculating key metrics, such as overall equipment effectiveness (OEE). In order to increase OEE, data generated by equipment in a production line is acquired and aggregated (preferably in automatically, see The Why of Data Collection).
Information is contextualized using business rules and user roles to create and maintain consistent functional and operational relationships between data elements from these disparate sources. For example, these products can demonstrate the relationship between allowed process variables and ranges of time series-based quality and yield data. It can also analyze information by using business rules to transform raw process data into meaningful KPIs. Data can also be filtered for any noise/outliers; visualized with a context-based navigation and drill-down capabilities; and presented or propagated to determine the factors and root cause disturbances that slow production or impact quality. Ultimately plant-level systems allow decisions to be made that will speed up throughput and increase first-run production.
How It Works
Configuring data sources for integration can be done through templates, which is analogous to selecting a printer for an office application. The real trick, however, is in having sound plant-level models, which are frameworks that portray accurate plant-level context and data management, within the application sets. In manufacturing, even small changes to a master plan can create a so-called "reality gap" and these are historically addressed by last-minute panicking and scrambling, all the while, the business protagonists are not always (if ever) conscious of the impact or even the validity of their "educated guess" decisions.
Thus, these new software applications make it possible to model the cascading consequences of anything users do in response to an unplanned event (like a customer doubling an order or a machine breaking down), which in turn, makes it possible to understand how the other, intertwined parts of the user organization and supply chain will be impacted by a change (see Bridging the Reality Gap Between Planning and Execution).
When users have information about unplanned events and how their responses will impact the company, they should have manufacturing intelligence that can guide them through the forking paths of exception-based, decision-making. The value of the plant-level information indeed changes when enterprises use it to support higher-level, strategic, and tactical business processes. For example, data generated for a department supervisor or for management purposes has one value, and the same data used for Sarbanes Oxley Act (SOX) compliance has another (see Attributes of Sarbanes-Oxley Tool Sets). Moreover, the value of quality assurance (QA) information increases substantially when used to support enterprise-wide warranty issues.
No comments:
Post a Comment