Reducing quality, energy & throughput losses with automated root cause analysis

Case study: Glass manufacturing

Download the case study

(5 min reading)

Reducing quality, energy & throughput losses with automated root cause analysis

Download the case study

(5 min reading)

Case study: Glass manufacturing

The Challenge:

Persistent losses in quality, throughput & energy efficiency

Like many glass production lines, this wine bottle production line suffered from a number of process-related losses.

These losses included:

Slide Melting related
defects
Size and shape
variability
Fabrication efficiency Production rate Quality variability And more...

Manual investigations by their process experts – including the use of a self-serve analytics platform – could not locate the root-cause of the problems. In most cases where losses occurred, no clear process inefficiencies were apparent, and the data tags they investigated were operating within the permitted ranges.

More fundamentally for the process experts was the question of overall, end-to-end process efficiency. Many of these KPIs appeared to be contradictory – e.g.  improving quality vs increasing throughput; or increasing throughput while reducing energy consumption. It didn’t seem possible to optimize one without harming the other.

At one of the first meetings with their plant leadership, the Seebo team unearthed a fascinating insight from their data over the past year of production: there were certain considerable periods – several days or even weeks – when their production line was consistently operating at much higher efficiency levels, across all of their KPIs! Then, for unknown reasons, efficiency levels would drop significantly for a period, before improving again.

The factory team were faced with the following question: how can they reduce process inefficiencies and increase the amount of time the line is operating at more optimal levels overall?


The Challenge:

Persistent losses in quality, throughput & energy efficiency

Like many glass production lines, this wine bottle production line suffered from a number of process-related losses.

These losses included:

Slide Melting related
defects
Size and shape
variability
Fabrication efficiency Production rate Quality variability And more...

Manual investigations by their process experts – including the use of a self-serve analytics platform – could not locate the root-cause of the problems. In most cases where losses occurred, no clear process inefficiencies were apparent, and the data tags they investigated were operating within the permitted ranges.

More fundamentally for the process experts was the question of overall, end-to-end process efficiency. Many of these KPIs appeared to be contradictory – e.g.  improving quality vs increasing throughput; or increasing throughput while reducing energy consumption. It didn’t seem possible to optimize one without harming the other.

At one of the first meetings with their plant leadership, the Seebo team unearthed a fascinating insight from their data over the past year of production: there were certain considerable periods – several days or even weeks – when their production line was consistently operating at much higher efficiency levels, across all of their KPIs! Then, for unknown reasons, efficiency levels would drop significantly for a period, before improving again.

The factory team were faced with the following question: how can they reduce process inefficiencies and increase the amount of time the line is operating at more optimal levels overall?


The Solution:

Identifying & preventing the hidden causes of production losses

The Seebo solution enables process experts to identify the hidden causes of their production losses, and gain clear recommendations as to how to prevent those process inefficiencies. 

Furthermore, production teams receive real-time alerts as soon as those inefficiencies are detected – enabling them to prevent losses before they occur.

The process inefficiencies Seebo identifies are often invisible to the human eye, in part because they are caused by a complex combination of behaviors among the hundreds or thousands of interrelated data tags on the line.

In this case, the Seebo solution discovered that whenever the bridgewall temperature was less than 1,560° Celsius, and at the same time the hot spot temperature was less than 1,580° Celsius, the number of melting-related defects increased very significantly.

This is a hugely important insight that the process experts could never have figured out on their own, since both of those tags remained within the permitted range. It was only the unique combination of those two specific ranges of tag values that was causing the losses.


The Solution:

Identifying & preventing the hidden causes of process-driven losses

The Seebo solution enables process experts to identify the hidden causes of their production losses, and gain clear recommendations as to how to prevent those process inefficiencies. 

Furthermore, production teams receive real-time alerts as soon as those inefficiencies are detected – enabling them to prevent losses before they occur.

The process inefficiencies Seebo identifies are often invisible to the human eye, in part because they are caused by a complex combination of behaviors among the hundreds or thousands of interrelated data tags on the line.

In this case, the Seebo solution discovered that whenever the bridgewall temperature was less than 1,560° Celsius, and at the same time the hot spot temperature was less than 1,580° Celsius, the number of melting-related defects increased very significantly.

This is a hugely important insight that the process experts could never have figured out on their own, since both of those tags remained within the permitted range. It was only the unique combination of those two specific ranges of tag values that was causing the losses.


The Result:

27%

Reduction in quality-related losses

2%

Increase in throughput

€1.42M

In savings & extra profits

The Result:

4.2% increase in clinker quality + 3.3% increase in kiln feed capacity =

$780K

extra profit on a single line

5.6% reduction in energy cost per ton of clinker = 

$521K

in energy savings on a single line

Greater kiln stability =

11.9%

lower NOx emissions

Get the full case study >

Download the case study

(5 min reading)

Get the full case study >

Download the case study

(5 min reading)