Implementation of machine learning and digital vision technologies

Projects / Implementation of machine learning and digital vision technologies

Implementation of machine learning and digital vision technologies. Implemented at the facilities of the enterprise LLC "Coşkunoz Alabuga".

Task: To implement the system of recognition for the presence of nuts on stamped parts and the manual stamping tracking system.

Customer: LLC "Coşkunoz Alabuga".

Implementation year: 2021

Recognition system for the presence of nuts on stamped parts

Solution

Developed a system that recognizes the presence of nuts being rolled into a part on automatic stamping line.
Ensured uninterrupted control of the final product, with a system of sound and visual notification of the operator.
Implemented collection and analysis of visual and temporal data of all controlled parts

Main components

Specialized cameras and sensors
Monitors
Information infrastructure for data collection and analysis

Result

Carried out 100% control without a human factor risk, data collection and analysis to help maintain tooling and equipment

Manual stamping tracking system

Solution

To eliminate the human factor, the entire sequence of stamped transitions at the manual stamping station is determining by deep machine learning
As a result of training, the system recognizes the sequence of stamped transitions without errors
In case of improper transitions laying by operator, the system stops the press to avoid damage of equipment, tooling and to eliminate possible accidents

Main components

Specialized cameras and motion sens
Monitors
Information infrastructure for data collection and analysis

Result

Created a system eliminating cases of improper laying of stamp transitions, which prevents accidents, as well as damage of equipment and tooling related with this.

View next project