• Written by Sonali Subhadarsini, Product Manager

The shift supervisor on the morning shift at a manufacturing plant is likely to be deciding this week how to allocate overtime, based on the current production schedule and the latest reports from the factory floor, which may not reflect the actual status of the production lines, as the reports can be delayed or incomplete, and the supervisor has to rely on their experience and knowledge of the plant's operations to make an informed decision.

The production questions that currently wait on a report

Many production questions in manufacturing currently wait on a report, which can be time-consuming to generate and may not provide the most up-to-date information, for example, the production manager may need to know the current output of a specific production line, or the quality control manager may need to know the number of defective products produced in the last shift, and these questions often have to wait for the daily or weekly report to be generated, which can cause delays in decision-making and response to production issues.

The production team may also need to know the status of maintenance activities, such as which machines are currently offline for maintenance, or when the next scheduled maintenance is due, and this information may not be readily available, or may be scattered across different systems and reports, making it difficult to get a clear picture of the current production status.

The data behind them: output against plan, cycle times, scrap and quality records, maintenance history

The data behind these production questions is typically stored in various systems, such as enterprise resource planning (ERP) systems, manufacturing execution systems (MES), and computerized maintenance management systems (CMMS), and includes information such as output against plan, cycle times, scrap and quality records, and maintenance history, which can be used to answer questions about production performance, quality, and maintenance, and to identify areas for improvement.

For example, a category manager may compare the output of different production lines to identify which lines are producing the most products, and which lines may be underperforming, or a quality control manager may analyze the scrap and quality records to identify trends and patterns in product defects, and to determine the root cause of quality issues, and this analysis can help to inform decisions about production planning, quality control, and maintenance scheduling.

Sector constraints: changeover losses and multi-plant comparability

Manufacturing plants often face specific sector constraints, such as changeover losses, which occur when production is stopped to change over to a different product, and multi-plant comparability, which can be challenging due to differences in production processes, equipment, and systems, and these constraints can make it difficult to compare performance across different plants, and to identify best practices and areas for improvement.

For instance, a production manager may need to minimize changeover losses by optimizing production schedules and workflows, or a plant manager may need to compare the performance of different plants to identify opportunities for improvement, and to share best practices across the organization, and this can be achieved by using data and analytics to inform decision-making, and to provide insights into production performance and quality.

What improvement looks like: OEE, unplanned downtime hours, scrap rate

Improvement in manufacturing production can be measured in various ways, such as overall equipment effectiveness (OEE), unplanned downtime hours, and scrap rate, which can provide insights into production performance, quality, and maintenance, and can help to identify areas for improvement, and to track progress over time.

For example, a production manager may aim to increase OEE by reducing downtime and improving cycle times, or a maintenance manager may aim to reduce unplanned downtime hours by improving maintenance scheduling and workflows, and these goals can be achieved by using data and analytics to inform decision-making, and to provide insights into production performance and quality, and by working with production teams to identify opportunities for improvement.

A reduction in scrap rate can also indicate an improvement in production quality, and can help to reduce waste and improve efficiency, and this can be achieved by analyzing scrap and quality records, and by identifying trends and patterns in product defects, and by working with quality control teams to implement improvements and to track progress over time.

What this does not do, or when it is the wrong choice

While AI for production in manufacturing can provide valuable insights and improvements, it is not a replacement for human judgment and expertise, and it is not suitable for all production decisions, such as those that require a high degree of creativity or nuance, and it is also not suitable for situations where the data is incomplete, inaccurate, or biased, and in these cases, other approaches may be more effective, such as working with subject matter experts to develop customized solutions.

Additionally, AI for production in manufacturing is not a one-size-fits-all solution, and it requires careful consideration of the specific needs and constraints of each plant, and it also requires ongoing maintenance and updates to ensure that the system remains accurate and effective, and this can be time-consuming and resource-intensive, and may require significant investment in IT infrastructure and personnel.