Global manufacturer of oral care products applies Digital Twins and Industrial AI to increase capacity and yield

Background

A global consumer goods manufacturer makes each of their best-selling oral care products across multiple global production sites. The company needed a way to identify recurring process inconsistencies that affected batch cycle time, batch quality, and yield across the different lines and operators.

In addition, a corporate initiative centered around productivity and efficiency created further interest in optimizing production processes, reducing material waste, and progressing towards the golden batch.

The Challenge

Production of each batch of product aligns to a recipe consisting of defined amounts for each ingredient and process steps. Despite having defined ingredients and standardized production processes, certain lines were challenged with longer cycle times, inconsistency in quality, and reduced output.

The company did not have enough visibility into their current process capabilities and equipment performance. The existing level of data analysis available to production and quality teams lacked the detail required to identify the source of problems affecting batch processes. This meant that they were unable to detect and understand why they were running above or below defined targets.

Despite operating within specified limits, the sites still believed there was opportunity for improvements through reducing variability and also optimizing processes. Through more granular understanding of their batch processes, they could perform a deeper analysis to gain insights and establish a path to improvement.

The Solution

The manufacturer teamed with TwinThread to implement a digital twin solution leveraging the Perfect Batch application, built to optimize batch processes with AI.

TwinThread seamlessly integrated into the manufacturers batch systems and equipment sensors at each site to collect and contextualize their data, and in a short time was able to establish a foundation for how each batch operates at each site.

Within the Perfect Batch application, each phase in the batch process is monitored for conformance to the ideal batch profile. Any deviation from the ideal control bands generates an anomaly that is automatically assessed for severity. This provides the ability for teams to visualize performance across lines, see performance statistics for each batch, and drill into batch details and the potential for improvement to address the highest variance or highest priority first.

The Perfect Batch application enabled quality, operations, and engineering teams to:

• Identify process parameters that can be adjusted to reduce cycle time, and understand the actual impact of cycle time
• Determine materials that were consistently added off target, although within capability limits
• Adjust material dosing and achieving target specification and consistency in quality
• Identifying inconsistent phase lengths that were driving longer batch cycle times
• Detect inefficiencies in individual equipment assets that required recalibration
• Compare similar recipes and processes across sites leading to collaboration and process improvements
• Enabled Center of Excellence (CoE) approach for a core team to monitor batch processes globally

The Result

Teams can continuously monitor batch process performance across sites and gain insights and automated actions leading to continuous improvements. In addition, analytics on the manufacturer’s equipment and processes serve as a foundation for their teams to collaborate around recipe and process optimization.

• Reduced batch cycle time
• Improved consistency in batch quality
• Reduced material loss
• Increased efficiency