Unlocking the Next Generation of Continuous Improvement

Over the past decade, productivity improvement has stagnated as many manufacturers reach a limit on the benefit of traditional continuous improvement.Now, as the shop floor becomes digital, a whole new world of data and insights are unlocking CI to reignite manufacturing productivity.
continuous improvement manufacturing

Table of Contents

Continuous improvement has always been part of manufacturers’ DNA. Methodologies such as TQC, Six Sigma, Lean, Theory of Constraints, and others have been used to build a culture of continuous improvement on the shop floor for many years. However, over the past decade, productivity improvement has stagnated as many manufacturers reach a limit on the benefit of traditional continuous improvement (CI).

Now, as the shop floor becomes digital, a whole new world of data and insights depend on unlocking CI applications to reignite manufacturing productivity.

continuous improvement

Source: U.S. Bureau of Labor Statistics

The Traditional Continuous Improvement Approach

Currently, most manufacturing experts rely on static point-in-time data sets that enable localized analysis of specific parts of the process, rather than systemwide. This includes defining a specific problem to be solved, measuring the magnitude of the problem, analyzing for root causes, understanding potential improvement levers, and controlling for sustenance of changes recommended. In Six Sigma parlance, this approach is known as DMAIC (Define, Measure, Analyze, Improve, and Control).

However, there are some significant challenges associated with this approach:

  1. Improvements coming out of these projects can take months, and, in many cases, the original problem may undergo significant change by the time improvements are made.
  2. Since root cause analysis is usually carried out based on a point-in-time snapshot of a localized process area data, the result is local optimization.
  3. For any systemwide analysis, most of the time is spent on combining different datasets and minimal time on actual analysis, resulting in a sub-optimal root cause analysis.
  4. Identifying improvements typically requires running multiple experiments (part of the Design of Experiments, or DOE, technique), which is both lengthy and repetitive.

Similar issues exist across all continuous improvement methodologies. As well, most organizations face the issue of needing highly skilled experts to lead the CI effort, where we are seeing a great deal of turnover.


Next-Gen Continuous Improvement Based on Real-Time, Systemwide Data

The answer to these challenges is to gain widespread access to real-time, systemwide data. Visibility across production lines, facilities, and even partner facilities significantly shortens data-gathering and baseline measurement, and assures accuracy up to the minute or hour. This in turn makes issue verification far more efficient. And, during the improvement identification phase, real-time data is combined with intelligent analysis tools to reduce the number of DOE experiments from high double-digit numbers to perhaps just a few.

All of this is made possible by advances in pipeline technologies which automate the process of data integration and complex modeling, creating a platform of data to support the next generation of CI. Coupling this manufacturing data platform with AI and machine learning algorithms is further transforming how data is turned into operational insights, providing:

  1. Access to a real-time stream of systemwide data made possible by new data pipeline technologies.
  2. The ability to see the effects of problems immediately, across the end-to-end line and even the supply chain. For example, on an auto assembly line, it is now possible for problems identified during final assembly to be correlated all the way back to the upstream step of forging, in real time.
  3. Democratization of data: Allows both deep CI-trained experts and line operators with no training in the CI toolkit to review system data, identify root causes, and make improvements. We have seen even untrained team members prove out hypotheses using real-time data and implement improvements


Continuous Improvement Based on Systemwide Analysis

Moving to systemwide analysis produces improvements of 3-5% from the moment the solution is activated, in our experience. In fact, in some cases, we’ve seen up to 10% productivity improvements over six to eight months, due to being able to run machines longer, run them faster, or upgrade the quality of the product.

This is a huge shift for manufacturers, one that will tremendously accelerate CI efforts and enable ongoing improvements to flourish as they did a decade ago. In this environment of continuous enhancement, manufacturers can oversee the entire production line as a single system, driving new levels of competitiveness and efficiency.

To learn more about how Sight Machine is helping manufacturers transform their continuous improvement efforts, visit www.sightmachine.com.

Picture of Sudhir Arni

Sudhir Arni

Senior Vice President, Business Outcomes at Sight Machine In this role, Sudhir leads a team of transformation leaders responsible for on-boarding all new customers and customer success managers who ensure adoption of Sight Machine technology across the global enterprise customer base enabling business expansion from existing customers

Curious about how we can help? 
Schedule a chat about your data and transformation needs.