Machine Learning

Transforming inventory counting to
Flexometal with Computer Vision and Machine Learning

Flexometal is a leading steel tube manufacturer with a 10-year history in the market. During this time, they have built a highly qualified team and are committed to meeting their customers' needs efficiently and quickly.

Problem

Despite its success in steel pipe production, Flexometal faced difficulties in inventory management. The manual process, which involved the use of plastic pipes, led to errors and delays in delivery times. Human errors were common, resulting in delays and the need to allocate more resources to ensure each order was delivered with the correct quantity of pipes on time.

Our Solution

When iNBest approached Flexometal, we understood their needs and conducted a thorough analysis of opportunities for improvement. We proposed an artificial intelligence-based solution utilizing computer vision and machine learning. This solution would overcome the challenge of inventory counting, reducing errors and significantly speeding up delivery times.

For the development of this solution, the following Amazon Web Services (AWS) were used:

Amazon Rekognition: Used for image recognition and analysis of steel tubes.

Amazon EC2: Used to host and run the Computer Vision application developed by iNBest on AWS infrastructure.

Factors determinants

This solution has allowed Flexometal to better control its inventories, which has been reflected in the company's Material Requirements Planning (MRP) system. The accuracy of the quantity of parts and tubes produced with a given material has enabled more reliable and accurate purchasing planning, leading to cost reduction.

With this solution based on computer vision and machine learning, Flexometal has been able to optimize its inventory counting, improve delivery time efficiency, and strengthen its inventory control. This has given the company a competitive advantage in the market, while allowing the team to focus on higher-value activities, such as product quality.

Results

We developed an innovative tool using Amazon Web Services (AWS) technologies. With this tool, simply capturing an image or photograph of the tubes, whether to be supplied or in the warehouse, makes it possible to determine the quantity of tubes in question with 99.9999% accuracy in a matter of seconds! After extensive testing, fine-tuning, and understanding Flexometal's products and processes, we implemented it.

We designed the tool with a high-quality interface, ensuring that not only qualified personnel could use it, but also existing Flexometal staff could manage it with brief training.

The results have been exceptional. Delivery times have been significantly reduced, both in hours and days. Furthermore, the team can now focus on higher-priority processes, such as product quality and continuous company improvement.

Contact us

and receive free consulting

Phone

(+52) 33 2309 0100
(+52) 55 6651 8800
+1 (973) 554 9068

Email

hola@inbest.cloud

Phone

(+52) 33 2309 0100
(+52) 55 6651 8800
+1 (973) 554 9068

Email

hola@inbest.cloud