At the St. Petersburg plant "Cvezy" the quality of bonding of plywood controls machine vision

2020

Continuation of the project of automation of quality control

At the St. Petersburg plant of Sveza quality control automation continues. The Sveza company reported about it on November 17, 2020. For determination of level and uniformity of putting glue on production the system of machine vision is implemented. A project objective is quality improvement of products and also operational efficiency due to cost reduction

For November, 2020 putting glue on sheets of an interline interval happens on valtsevy installations in the automatic mode. Computer control will provide the continuity of process and stability of result, will exclude excess of glue on sheet surfaces and also its insufficient quantity which leads to formation of defects - "bubbles". Accurate following to the program will increase durability of gluing together of the plywood sheet, will reduce defects percent.

For an exception of low-quality filler on roller installation the "machine vision" system ("computer vision") is implemented. The essence is that after putting glue sheets of an interline interval pass under the video camera which fixes uniformity and density of filler. Evenly put layer has a certain color and tonality. In case of a deviation from the set parameters, "machine vision" reports about it to the operator of installation. The employee removes not missed the mark sheet of an interline interval and works on elimination of the reason.

Investments into the project made more than 1.7 million rubles. The expected economic effect of implementation of an innovation is not less than 800 thousand rubles a year. The idea was developed by specialists of the Center of technology development of Sveza.

At the St. Petersburg plant there is a multi-layer system of quality control. If we do not miss the mark an interline interval, then we, certainly, will detect the formed emptiness in the ready sheet of plywood: for this purpose we have an ultrasonic defectoscope. But it will mean that the sheet of plywood or will go to defects, or it will reduce grade. Prevent such losses to us and "machine vision" will help.

The center of technical development began to work at the St. Petersburg plant in 2019. Its key task consists in search and implementation of the innovative solutions for the purpose of increase in efficiency of technology process and cost reduction.

Tracking of defects by means of high-precision video cameras and neural networks

At Sveza plant in Manturov project implementation, the product quality improvement connected with system implementation continues. Tracking of defects will happen by means of high-precision video cameras and neural networks. It became known on June 11, 2020.

The operation principle of the Machine Vision or Computer vision project consists that ready sheets of plywood pass on the conveyer belt at great speed under special video cameras. Cameras "take" images and transfer data to a system into which thousands of photos of possible defects are loaded previously. The "machine vision" working by the principle of a neuronet in read seconds defines products sort. At the same time, if a system detects defects, then it transmits a signal to the pipeline and the sheet "goes" to a separate pocket.

Importance of implementation of Machine vision is caused by several reasons. First, it allows to eliminate a human factor in the course of sorting of plywood and to unify grade of finished goods on all branches of Sveza. Secondly, will allow to accelerate sorting process. For June 10, 2020 at manturovsky plant to define a sort of the ready sheet of plywood, 16 seconds are required on average. "Computer vision" will allow to accelerate process several times.

It is absolutely uncommon task: the quantity and options of shortcomings of wood are huge. Some defects are well visible to "machine vision" in the h \color, others can be revealed only the color camera, some in red light, - Evgeny Alexandrov, the head of service on technology and quality of Sveza plant in Manturov notes.

On Svez the similar project is implemented for the first time. It is remarkable that as the pilot platform the manturovsky plant was selected. Exactly here the neuronet is trained to sort plywood: more than 30,000 sheets are finished already shooting. In a system more than 700 photos of different defects are brought. Purchased and ready to operation of the video camera. In the fall tests are planned. On the basis of the carried-out analysis the decision on implementation of Machine vision at other plants of the company will be made.

Installation of hygrometers and multichannel defectoscopes

On April 9, 2020 it became known that within a comprehensive investment program on quality control on production lines of Sveza company hygrometers and multichannel defectoscopes are installed.

In the conditions of a pandemic and economic downturn product quality becomes one of the main advantages of production companies. Plants of Sveza are equipped with advanced technologies and systems for maintenance of stable product quality.

Within a comprehensive investment program completing came to the end with multichannel defectoscopes of all lines of sorting of white plywood. For Russia this special equipment. Svez - one of the few producers of plywood which uses it. The device helps to prevent revenues to the market of products with internal defects and to do plywood such what expect it to receive clients. Plywood passes through the defectoscope before a final stage of sorting. The device illuminates the sheet ultrasound: if it passes without obstacles, means emptiness is not present if "sees" a vacuum, then sprays red paint, rejecting the sheet. The equipment allows to control the hidden defects, such as stick also weak bonding. It is one more step on strategy implementation of the company on Customer Centricity which means continuous improvement of quality and delivery only of high-quality products.

One more program for products quality excellence - the project on detection of defects by means of machine vision. For April, 2020 specialists "train" a system, using base from thousands of photos of defects that it could distribute unmistakably sheets of plywood on sorts. The pilot platform selected plant from Manturov of the Kostroma region. Exactly there the equipment for the analysis was installed and sorting of 10,000 sheets of plywood is finished shooting. It is a basis for future network functioning. Selection of cameras and lighting for a possibility of identification of all defects of plywood is at the same time conducted.

It is an uncommon task: the quantity and options of shortcomings of wood are huge. Some defects are well visible to "machine" in the h \color, others can be revealed only the color camera, some in red light, speaks Evgeny Alexandrov, the head of service on technology and quality of Sveza plant in Manturov

All plants are brought to the uniform unified quality management system. In structure all moments are considered: multistage control, system approach to problem solving and methodology 8D for identification of root shortcomings. High quality of products of Sveza - to a large extent also result of permanent communication and open dialog with clients. We receive a feedback from our buyers and we adjust work depending on wishes", noted Oleg Chistyakov, the operating officer of Sveza company