Knauf is a company where manufacturing is king with a highly efficient manufacturing process. Based on their mission they are committed to improving their products, the production experience for their employees, and their impact on the environment.
The Problem
Knauf is split into 2 main business segments: Drywall (gypsum) and Insulation (rock & glass mineral wool). The production environments are dangerous, therefore quality is not being assessed in real time, leading to safety issues, excess waste, and lower-quality products.
Drywall: The drywall manufacturing line currently has employees reviewing the health of the line after the product has been laid and heated. Knauf is looking to conduct product inspections remotely as the current environment is unsafe for humans due to high levels of noise and air particles. One main concern is that air bubbles can form within the drywall during production which has a significant impact on quality. Currently, quality inspections cannot be done in real-time and are conducted via spot checking only once per 8-hour shift.
Insulation: Insulation products are developed by dropping molten glass or rock onto spinner wheels. The melt stream of rock and glass has certain variable properties that will greatly affect the quality of the end product that cannot be measured by human inspection or current non-vision measurement systems. Waiting to inspect the quality until after production process is complete leads to excess waste and lost time.
Key Challenges
Challenge 1: Extreme Environments
For both insulation and gypsum, the operating environment can be extremely dusty, noisy, and hot. Some equipment may vibrate during production. Any vision system that is installed to monitor production must be able to withstand extreme operating conditions.
Challenge 2: Real-Time Analysis
In each of Knauf’s use-cases, real-time analysis is needed. For Vision AI and Closed-Loop Manufacturing solutions, this includes conducting inferences at or near real-time video speeds of 30FPS (frames per second). This data must be logged to the control systems within milliseconds, accurately analyzed, and feedback/recommendations to fix any quality variation issues immediately.
Challenge 3: Mission Criticality
The vision systems will be directly tied to the control systems of the manufacturing lines. As such, these are considered to be mission-critical components of the manufacturing line which will need to be operationally sound, extremely accurate, and have high uptime. Any downtime or lack of reliability will have significant consequences on production.
The Solution
ClearVision @ The Edge & Vision System Pipeline
ClearObject provided Knauf with a highly available and reliable vision system with real-time deep learning prediction capabilities that will allow users to monitor results through an edge vision streaming and vision dashboard viewable from anywhere.
The ClearVision dashboard and notification triggers allow Knauf to monitor and be alerted if any of their manufacturing lines deviate from standards. With near real-time notifications issues can be stopped and corrected before they become actual problems.
Implementation
In just months, ClearObject collaborated with Knauf to develop and deploy a solution designed for their specific needs. Beginning with a Vision AI Workshop, opportunities were identified and explored, leading to an Opportunity Assessment with value and effort estimates to fully deploy in production. Once defined, ClearObject moved to creating and training a purpose-built Vision AI Model Proof of Concept and deployed it to Knauf’s edge hardware. Once the concept was fully validated, the teams were well-prepared to expand the roll-out.
“The ClearObject team came in and really listened to our needs. Together we collaborated and figured out solutions for both our Drywall and Insulation processes that would make a real impact. They knew exactly what we needed to bring this vision to reality.” – Trent Miller, Head of AI and Data Science 
Impact
Since launching the systems in late 2022, Knauf has seem immediate and dramatic impacts for both the bottom line and towards reaching their ESG (environmental, social and governance) goals. In the first year after launch, Knauf is predicting millions in savings with an ROI of 511%.
Key Wins
- 50% reduction in CO2 emissions per implementation, helping Knauf reach ESG goals and providing positive tax incentives
- Up to 5% scrap rate reduction leading to significant cost savings
- Measurable improvements in worker safety
- 0.5% increase in efficiency while increasing quality
“We are thrilled with the results from the ClearVision @ the Edge solution. Our scrap rates and overall environmental impact started dropping on day one, and that has a real impact on the bottom line. From the sales process to the Vision AI workshop and beyond, the ClearObject team was easy to work with and delivered on their promises.”
– Trent Miller, Head of AI and Data Science
The Team
A full team of ClearObject Vision AI experts partnered and collaborated with the Knauf team to build this bespoke solution. ClearObject’s team consisted of Product Management, Project Management, Data Scientists, and multiple developers who worked closely with key Knauf stakeholders to understand their business goals and design a system that would achieve them.
Download the full case study
