• Written by Sonali Subhadarsini, Product Manager

On the production floor, the shift supervisor checks the maintenance log to see if any equipment is due for scheduled downtime, while the category manager compares two suppliers to determine which one can deliver the necessary parts on time. In this complex environment, introducing a new system like AsscherAi, which allows employees to query production data in natural language, requires careful consideration of governance and risk.

Enforcing permission at retrieval

One common mistake is to try to instruct the model on what data it can and cannot return, but this approach is flawed because it relies on the model's understanding of the instructions, which can be imperfect. Instead, permissions should be enforced at the point of data retrieval, so that even if the model returns sensitive information, it will be filtered out before it reaches the user. This approach ensures that access to data is controlled at the point of use, rather than relying on the model to make decisions about what data to return.

This requires a deep understanding of the organisation's data architecture and the different roles that employees play in the production process. For example, the shift supervisor may need to see maintenance schedules, but not sensitive information about supplier contracts. By enforcing permissions at retrieval, the organisation can ensure that each employee only sees the data they need to do their job, and that sensitive information is protected.

Auditability

Auditability is critical in a production environment, where decisions can have significant consequences for the organisation. When using AsscherAi, it is essential to have a clear record of what was asked, what was returned, and on what data the response was based. This allows the organisation to track how data is being used and to identify any potential issues or biases in the system.

This is particularly important in a production environment, where data is often used to make decisions about equipment maintenance, supplier contracts, and inventory management. By having a clear audit trail, the organisation can ensure that decisions are being made based on accurate and up-to-date information, and that any errors or inconsistencies can be quickly identified and corrected.

Production data

The effectiveness of AsscherAi in a production environment depends on the quality and accuracy of the underlying data. This includes output against plan, cycle times, scrap and quality records, and maintenance history. By analysing this data, the organisation can identify trends and patterns that can inform decisions about production planning, inventory management, and equipment maintenance.

For example, the category manager may use AsscherAi to compare the performance of different suppliers, based on data about delivery times, quality, and cost. By analysing this data, the category manager can make informed decisions about which suppliers to use, and how to negotiate contracts that meet the organisation's needs. The production team can also use AsscherAi to track output against plan, and to identify areas where production is falling behind schedule.

Unplanned downtime and stock cover

One of the key benefits of using AsscherAi in a production environment is the ability to reduce unplanned downtime and improve stock cover. By analysing data about equipment maintenance, production schedules, and inventory levels, the organisation can identify potential issues before they occur, and take proactive steps to prevent them.

For example, the shift supervisor may use AsscherAi to identify equipment that is due for maintenance, and to schedule downtime accordingly. This can help to prevent unexpected equipment failures, which can cause costly delays and disrupt production. The organisation can also use AsscherAi to track stock levels and to identify areas where inventory is running low, allowing them to take proactive steps to replenish stock and prevent shortages.

By reducing unplanned downtime and improving stock cover, the organisation can improve overall efficiency and reduce costs. This can be achieved by optimising production planning and scheduling, and by using data to inform decisions about inventory management and equipment maintenance.

Measuring performance

The effectiveness of AsscherAi in a production environment can be measured using a range of key performance indicators (KPIs), including Overall Equipment Effectiveness (OEE), unplanned downtime hours, and scrap rate. By tracking these KPIs, the organisation can see how AsscherAi is impacting production, and make adjustments as needed to optimise performance.

For example, the production team may use AsscherAi to track OEE, and to identify areas where equipment is not running at optimal levels. By analysing this data, the team can identify opportunities to improve equipment performance, and take steps to implement changes that will improve overall efficiency. The organisation can also use AsscherAi to track unplanned downtime hours, and to identify areas where equipment maintenance is causing delays.

Limitations

While AsscherAi can be a powerful tool in a production environment, it is not a solution for every problem. In some cases, the organisation may need to use other tools or approaches to address specific challenges or issues. For example, AsscherAi may not be able to provide real-time data about equipment performance, or to track complex production schedules.

In these cases, the organisation may need to use other systems or tools to supplement AsscherAi, or to provide additional functionality. The organisation can contact us to discuss their specific needs and to determine the best approach for their production environment.