August 27, 2015 at 10:04Industrial video surveillance: we remove technological data from cameras and predict emergency situations
Very far beyond the Moscow Ring Road, where only geologists and rare bears roam (looking ahead - they are also toxic addicts), there is a production complex. Mined ore arrives at the entrance, and concentrate at the exit, which may later become stagnant.
If, under certain conditions, a worker enters this zone, his family will be paid insurance.
At this plant, of course, a bunch of sensors are reported that report accidents. Nevertheless, a second confirmation loop was needed, because before the start of the project, a person had to resort to the workshop or point and visually confirm the situation. Before that, it was impossible to take any critical measures.
This is a color frame from one of the cameras. Everything, everything is covered in gray dust.
Unlike warm and attractive offices and sterile data centers, electromagnetic radiation from cables with a foot thick and noise from the start-up of various equipment is constantly sick at this facility. And yet there is excellent dust, which accumulates a centimeter per day.
So, there is a set of sensors at the factory, which signals the dispatcher to emergency situations. Since the sensors very often behave strangely, plus they give an incomplete picture of the scale of the situation and the nature of its development in general, to make a serious decision you need to send a person to see what happens. The intercom dispatcher calls the nachce, who is already looking for the foreman and asks to figure it out. There are two answers: “No, the computer drives” and, much less often, “Mikhalych, here ####, come on #### urgently!”. A few minutes later comes a verbal description of the accident. If the accident is serious - in a very concise, capacious and figurative form. It is this form (4-5 minutes late) that causes the conveyor to stop, an emergency pressure relief on a number of units, breakdown of expensive equipment during an emergency shutdown, and other things,
Of course, it was impossible to put up with financial losses.
Initially, the customer turned to us to calculate the cost of the machine vision system, which allows automatic tracking of all problem points. A list of nodes and a list of possible accidents was attached. For example, in this vat nothing should overflow; this node should not work while people are nearby; sometimes bears come to the pipeline out of curiosity; here there is a breach in the pipeline, and here it is necessary to make sure that a large piece, garbage, or the tape does not tear on the conveyor suddenly.
Running a little ahead, I will show a frame from a line where, for example, you need to control passing crowbars, sticks and other large-sized items. The size of the fraction is still controlled in the West, but this is already much more expensive because of the requirements for picture, lighting, and because of the features of the code:
For such systems, there are ready-made industrial solutions developed for American, German and Japanese industries. Their main problem is that each non-standard detector requires a serious amount of code and recognition tests. We spoke with both domestic and Western developers of such systems, and calculated the first budget. The main problems are these non-standard detectors and plots that are difficult to recognize as above, with almost no contrast.
The budget turned out to be very large, so the customer went to think further. According to his calculations, it turned out that it would be much cheaper in the 10-year perspective to raise the control room and put people there who know the production like the back of their hand and are perfectly able to recognize emergency situations. Plus fasten basic automation like focus in the picture on triggering the sensor, basic tracking of movement in areas where there should be no movement, etc. In general, a person should be - like a grandmother on an escalator in the subway.
So, we assembled a system based on cameras, simple sensors, and in the software part — basic (library) detectors.
Correctly, the system is called "technological vision." The technologist remotely watches what is happening at his factory there and does not enter into an aggressive environment. But in this aggressive environment of the plant are our cameras. The concept of technological vision also includes such solutions as specialized analytics for the site. For example, if such a machine is installed, then the task is to mix the components and get a certain type of foam at the output. Machine vision determines the size of the foam cell - depending on the size of the cell, the analyst determines which reagent to add and immediately issues a command to the machine controller. We have not implemented this yet, but, in general, it is quite real. Another thing is that, I repeat, it is here, due to special conditions, that a living technologist is more economical.
The first problem of the project was that it was necessary to draw up a specific project proposal. If on a regular project this can be done in a week of sitting with plans and models and one trip, then this did not work out here. We went there several times. We walked around each workshop, talked with technologists, asked everyone what they needed to see and how they would like the sensors to be located.
Here is a picture on the phone from the survey and the outlined field of view of the camera. The
dimensions of the plant are enormous (tens of kilometers), there aren’t exact plans everywhere, there’s no way to attach to something in place to make a plan for installers in the future. As a result, we simply filmed certain places and marked the points under the cameras and another weak point on them. Like that:
The point is to remove not only the montage point, but also immediately with the technologist what the camera will see. To do without misunderstandings.
We did this for about two weeks. Usually, design takes so much, but it took 2 months to complete the project. I must say that I walked around the object with an office laptop, which during that time fell a couple of times, crashed, and even six months after the system was commissioned, there were still traces of local dust on it.
Switching itself was relatively simple. The complex has an optics ring with branches, so there are no problems with the channel for the HD picture. Optics is there, by the way, because copper simply will not work. Long crossings between workshops are only possible on existing cable racks. And there are power cables with a cross section of a few centimeters. And there were not a dozen of them living there. Copper was only from the shields to the cameras themselves, and in the cable with several levels of shielding.
The panels were assembled by themselves - they calculated the temperature regimes and put industrial noise-proof switches inside. As practice has shown, this works great, and does not become a bottleneck in any way. The problem was that there were no cameras as such for very hard industrial use. In the sense that they still had the appropriate certificates. Therefore, again, I had to take it in special covers, make grounding, screw on special “wipers” from above to shake off the dust. One large vendor found the casing with the wipers very well and immediately got up as it should.
Of the 165 cameras at the commissioning site, 19 fell off immediately. Transportation of this or factory defects is unknown. It was necessary to check in Moscow after arrival, but the customer needed as soon as possible, so we sent immediately to the complex, without unpacking.
When they mounted their tracks, they learned what legacy isolation is. The plant has many lines that have not been used since the days of the USSR, because they are duplicated, replaced and not unmounted, etc. In most cases, the customer gave us trays, and we mounted them. But on a number of runs there were consoles with large distances between the supports. A thick self-supporting cable felt great there, and everything else was attached, in fact, to it. So, after 20 years of operation in an aggressive environment, the insulation on such cables turns into dust. While you are standing and watching, everything is fine. You touch the braid - it pours with a quiet rustle.
In one of the legacy workshops
Previously, a pipeline break could erode the embankment at the nearest lake in 15-20 minutes. Now the blur time has not changed, but the reaction time has changed. Visual confirmation can be obtained immediately and immediately make a decision. Previously, this required jumping into an all-terrain vehicle and heading to the other end of the territory, for example. Here it is also necessary to add that it was not scary to walk along this lake near the complex only during both weeks of summer. The rest of the time, toxic-bear bears have hot parties there. These animals adore strong unusual smells - they are like bears reading for IT specialists in bears. Therefore, checking the sensor every time was somehow not very interesting. We were also told stories about wolves, but we did not see them.
In addition to reducing the reaction time, the downtime after the accident also decreased. Before, after the stop, the big problem was to make sure that all the repairmen left the unit. This was done along the same lines and with difficulty. Now there is still visual confirmation.
If, under certain conditions, a worker enters this zone, his family will be paid insurance.
At this plant, of course, a bunch of sensors are reported that report accidents. Nevertheless, a second confirmation loop was needed, because before the start of the project, a person had to resort to the workshop or point and visually confirm the situation. Before that, it was impossible to take any critical measures.
This is a color frame from one of the cameras. Everything, everything is covered in gray dust.
Unlike warm and attractive offices and sterile data centers, electromagnetic radiation from cables with a foot thick and noise from the start-up of various equipment is constantly sick at this facility. And yet there is excellent dust, which accumulates a centimeter per day.
Formulation of the problem
So, there is a set of sensors at the factory, which signals the dispatcher to emergency situations. Since the sensors very often behave strangely, plus they give an incomplete picture of the scale of the situation and the nature of its development in general, to make a serious decision you need to send a person to see what happens. The intercom dispatcher calls the nachce, who is already looking for the foreman and asks to figure it out. There are two answers: “No, the computer drives” and, much less often, “Mikhalych, here ####, come on #### urgently!”. A few minutes later comes a verbal description of the accident. If the accident is serious - in a very concise, capacious and figurative form. It is this form (4-5 minutes late) that causes the conveyor to stop, an emergency pressure relief on a number of units, breakdown of expensive equipment during an emergency shutdown, and other things,
Of course, it was impossible to put up with financial losses.
Initially, the customer turned to us to calculate the cost of the machine vision system, which allows automatic tracking of all problem points. A list of nodes and a list of possible accidents was attached. For example, in this vat nothing should overflow; this node should not work while people are nearby; sometimes bears come to the pipeline out of curiosity; here there is a breach in the pipeline, and here it is necessary to make sure that a large piece, garbage, or the tape does not tear on the conveyor suddenly.
Running a little ahead, I will show a frame from a line where, for example, you need to control passing crowbars, sticks and other large-sized items. The size of the fraction is still controlled in the West, but this is already much more expensive because of the requirements for picture, lighting, and because of the features of the code:
For such systems, there are ready-made industrial solutions developed for American, German and Japanese industries. Their main problem is that each non-standard detector requires a serious amount of code and recognition tests. We spoke with both domestic and Western developers of such systems, and calculated the first budget. The main problems are these non-standard detectors and plots that are difficult to recognize as above, with almost no contrast.
The budget turned out to be very large, so the customer went to think further. According to his calculations, it turned out that it would be much cheaper in the 10-year perspective to raise the control room and put people there who know the production like the back of their hand and are perfectly able to recognize emergency situations. Plus fasten basic automation like focus in the picture on triggering the sensor, basic tracking of movement in areas where there should be no movement, etc. In general, a person should be - like a grandmother on an escalator in the subway.
So, we assembled a system based on cameras, simple sensors, and in the software part — basic (library) detectors.
Technological vision
Correctly, the system is called "technological vision." The technologist remotely watches what is happening at his factory there and does not enter into an aggressive environment. But in this aggressive environment of the plant are our cameras. The concept of technological vision also includes such solutions as specialized analytics for the site. For example, if such a machine is installed, then the task is to mix the components and get a certain type of foam at the output. Machine vision determines the size of the foam cell - depending on the size of the cell, the analyst determines which reagent to add and immediately issues a command to the machine controller. We have not implemented this yet, but, in general, it is quite real. Another thing is that, I repeat, it is here, due to special conditions, that a living technologist is more economical.
The first problem of the project was that it was necessary to draw up a specific project proposal. If on a regular project this can be done in a week of sitting with plans and models and one trip, then this did not work out here. We went there several times. We walked around each workshop, talked with technologists, asked everyone what they needed to see and how they would like the sensors to be located.
Here is a picture on the phone from the survey and the outlined field of view of the camera. The
dimensions of the plant are enormous (tens of kilometers), there aren’t exact plans everywhere, there’s no way to attach to something in place to make a plan for installers in the future. As a result, we simply filmed certain places and marked the points under the cameras and another weak point on them. Like that:
The point is to remove not only the montage point, but also immediately with the technologist what the camera will see. To do without misunderstandings.
We did this for about two weeks. Usually, design takes so much, but it took 2 months to complete the project. I must say that I walked around the object with an office laptop, which during that time fell a couple of times, crashed, and even six months after the system was commissioned, there were still traces of local dust on it.
Switching itself was relatively simple. The complex has an optics ring with branches, so there are no problems with the channel for the HD picture. Optics is there, by the way, because copper simply will not work. Long crossings between workshops are only possible on existing cable racks. And there are power cables with a cross section of a few centimeters. And there were not a dozen of them living there. Copper was only from the shields to the cameras themselves, and in the cable with several levels of shielding.
The panels were assembled by themselves - they calculated the temperature regimes and put industrial noise-proof switches inside. As practice has shown, this works great, and does not become a bottleneck in any way. The problem was that there were no cameras as such for very hard industrial use. In the sense that they still had the appropriate certificates. Therefore, again, I had to take it in special covers, make grounding, screw on special “wipers” from above to shake off the dust. One large vendor found the casing with the wipers very well and immediately got up as it should.
Of the 165 cameras at the commissioning site, 19 fell off immediately. Transportation of this or factory defects is unknown. It was necessary to check in Moscow after arrival, but the customer needed as soon as possible, so we sent immediately to the complex, without unpacking.
When they mounted their tracks, they learned what legacy isolation is. The plant has many lines that have not been used since the days of the USSR, because they are duplicated, replaced and not unmounted, etc. In most cases, the customer gave us trays, and we mounted them. But on a number of runs there were consoles with large distances between the supports. A thick self-supporting cable felt great there, and everything else was attached, in fact, to it. So, after 20 years of operation in an aggressive environment, the insulation on such cables turns into dust. While you are standing and watching, everything is fine. You touch the braid - it pours with a quiet rustle.
In one of the legacy workshops
Total
Previously, a pipeline break could erode the embankment at the nearest lake in 15-20 minutes. Now the blur time has not changed, but the reaction time has changed. Visual confirmation can be obtained immediately and immediately make a decision. Previously, this required jumping into an all-terrain vehicle and heading to the other end of the territory, for example. Here it is also necessary to add that it was not scary to walk along this lake near the complex only during both weeks of summer. The rest of the time, toxic-bear bears have hot parties there. These animals adore strong unusual smells - they are like bears reading for IT specialists in bears. Therefore, checking the sensor every time was somehow not very interesting. We were also told stories about wolves, but we did not see them.
In addition to reducing the reaction time, the downtime after the accident also decreased. Before, after the stop, the big problem was to make sure that all the repairmen left the unit. This was done along the same lines and with difficulty. Now there is still visual confirmation.
References:
- Video surveillance systems for data centers: who touched the rack and why
- Number Recognition (2012)
- Face recognition for access control systems
- Mail, if suddenly there are questions not for comments - DNikiforov@croc.ru