About us:

Constems-AI Computer Vision technology for quality checks in manufacturing lines: eliminating inefficient ad hoc human checks in manufacturing lines and lowering product rejection by end users due to quality gaps. Our AI-based visual inspection CAInatics platform is a simple to deploy system on manufacturing lines that provides production data and real-time OEE reports to QC teams. The CAInatics platform employs cloud and edge-based systems to build an architecture that includes visual inspection equipment with real-time backend support for AI models that detect any quality lapses in real-time.

The Client:

A leading agricultural manufacturing company with operations in over 90 countries worldwide. They are pioneers in the production of FMCG/Agri-Based products. The company has over 100 years of experience and over 150 million satisfied customers worldwide.

The Problem:

The company was having a major problem with agri-based product inspection. Previously, they relied heavily on human inspection and rejection of defects for their products. Unfortunately, it was an expensive and inefficient transaction. Because very minute substances must be inspected, human-based quality checks are ineffective and time-consuming. This resulted in a significant loss of efficiency and end-product quality. The company discovered that nearly 90% of the raw materials were ignored, while quality issues were reported in more than 30% of the finished products. The QC team was in a terrible situation.

Mr. A.S Tripathi (Director, Quality Check Team) was in a lot of trouble as a result of the loss, and the entire team was looking for alternatives. They have deployed a few teams in the market to investigate potential alternatives to ad-hoc human inspection.

The Solution:

The research team contacted Constems-AI, whose core business is an AI-based vision inspection system. Our team went through the entire process in the factory to get a better understanding of the situation. Following a thorough examination of the workflow, the Constems-AI product deployment team devised a solution to install our AI-based inspection/rejection system "FME-1000."

The AI-based Quality inspection product "Foreign Material Eliminator-1000 (FME-1000)" is unique in its class and is part of the Industry 4.0 initiative. Constems-proprietary AI's CAInatics, an AI vision-based inspection platform, aids in the detection of any foreign material above 1000 microns (1 mm) on the production line, with the ability to integrate with BI Tools and RPAs for rejections of any such foreign materials to ensure the highest quality of agri and processed food products. It improves "Quality Inspection" capabilities to more than 98 percent, according to SLA (Service Level Agreement), and thus improves the quality of food / agri-product to end consumers.

It also includes a root cause analysis tool that employs an intelligent ML algorithm to comprehend patterns in current processes and identify gaps. It also provides defect image data that is time-stamped for easy retrieval and auditing. The FME-1000 is intended to automatically escalate any spike in quality issues and clearly identify the quality gap. It enables prompt action based on the concerns raised. FME-1000 also provides remote access to Online Equipment Efficiencies (OEE) key performance indicators, time-series data for quality control, and the performance status of the production line via our proprietary CAlnatics platform. 

The Impact:

We were able to deploy our FME-1000 at the aforementioned company in less than 35 days. The installation proved to be a boon for the QC team, as they were able to inspect raw materials up to 100% while involving very few human resources, resulting in a savings of more than 25% in human resource expenses. Furthermore, the team reported a massive reduction of more than 40% in quality issue cases.

The QC team now has complete control over the raw material quality inspection and can manage the process from their offices.