Introduction

Daily production is a key performance indicator (KPI) in continuous processes, helping process experts track efficiency and performance. However, in many cases, daily production is not readily available as a default variable in data historians. A recent question in the TrendMiner Community addressed this challenge: How to calculate a daily average of a signal with a single result per day?. This article explains four different approaches to calculating daily production in TrendMiner, depending on the user’s needs.

Continuous Rolling Last 24 Hours Production

Background

This method ensures that at any given time, the production of the last 24 hours is always available as a continuously updating value. This is particularly useful for real-time monitoring and trend analysis.

Steps

1. Load the relevant tag related to production.

2. Set the context and focus chart to visualize the required time range.

3. Create a new aggregation:

   • Select the production signal.

   • Use a backward aggregation over 24 hours.

   • Apply the integral operator to calculate the total production.

   • Define the appropriate unit and save the new variable.

This approach provides a continuously updated rolling production metric, ideal for evaluating short-term performance.

Daily Production as Peaks

Background

This method calculates daily production by capturing a single peak value at the beginning of each day. It allows for a simplified overview of production performance with distinct daily markers.

Steps

1. Start from the continuous rolling production calculation.

2. Create a new formula to display only the daily peak:

   • Identify when a new day begins.

   • Assign the daily production value at that moment.

   • Keep all other values at zero.

3. Save the new variable to enable easy tracking of production peaks.

This method is particularly useful for reporting and detecting daily variations in production output.

Linear Interpolated Daily Production

Background

This method calculates a single production value per day while interpolating between data points, offering a smooth daily trend for process analysis.

Steps

1. Start from the previous peak-based production calculation.

2. Modify the formula to interpolate between data points:

   • Use a condition to determine when a new day starts.

   • Retain the production value for that day.

   • Apply interpolation for a smooth trend between data points using sqrt(-1). This mathematical trick forces TrendMiner to interpolate between the recorded data points, ensuring a continuous trend.

By interpolating between values, this method provides a clearer visualization of production trends over time.

Stepped Daily Production

Background

This approach provides a constant daily production value for each day, ensuring a stepwise representation of daily production without interpolation.

Steps

1. Apply an aggregation:

   • Use an integral calculation over 24 hours.

   • Set the aggregation direction to forward.

   • Save the new variable to track daily production with stepwise consistency.

2. Define a formula that assigns a constant value for each day:

   • If the current day differs from the previous day, assign the aggregated daily production value.

   • If the current day is the same as the previous day, maintain the existing value.

   • If neither condition is met, apply sqrt(-1) as a trick to interpolate correctly between steps.

This method ensures that the daily production remains constant throughout the day before updating at the next daily boundary.

Conclusion

Each method serves different analytical needs:

• Rolling 24h production → Ideal for real-time monitoring.

• Daily production as peaks → Best for capturing distinct daily values.

• Linear interpolated production → Useful for visual trend analysis.

• Stepped daily production → Suitable for stepwise reporting.

By leveraging these approaches, TrendMiner users can efficiently monitor and analyze their daily production performance.

Do you have experience with these methods? Share your insights and use cases in the comments!