Gazprom Neft (Vizorlabs Platform Video Analytics Solution)

2020

Digital mass and size measurement of MTR in stock

By order of PJSC Gazprom Neft, the company VizorLabs developed and tested a prototype software and hardware complex for automatic weight and size measurement and recognition of MTR (spare parts, consumables and other products) in the warehouse using computer vision and artificial intelligence. The technology allows you to recognize single large parts, and parts in a group package on a pallet.

The complex is integrated with all WMS systems known as of February 2021 and allows you to automate the process of acceptance/issue of MTR and ordering transport for the dimensions of MTR.

The solved problem

The traditional acceptance/issue technology of MTP involves the manual entry of information in WMS or the maintenance of primary accounting in paper journals and the transfer of data to WMS with a delay in time, which reduces the additional labor and time costs, reduces the efficiency of accounting, leads to errors when entering the MTP item in WMS. Mass and size characteristics of MTP may not be taken into account.

Decision

Automatic measurement and determination of the MTR nomenclature using a software-hardware measuring complex with computer vision and artificial intelligence technology. The use of this technology allows you to reduce the labor costs of personnel as part of a comprehensive automation of warehouse storage.

The complex includes 3 static video cameras, software, server video analysts as well as industrial electronic scales. Cameras are mounted on a fixed ramp above the scales in such a way that their fields of view cover the measured object from all sides and intersect with each other (right, left, top).

Dual infrared and sliding shutter cameras are used, which allows you to measure the distance to the surfaces of an object and compare this data with the image with a regular RGB camera to calculate the linear dimensions of the object. The distance from the camera to the object does not exceed 3 m and must be proportional to the dimensions of the measured objects.

When placing the MTP on the weight platform, the complex receives information on the weight of the part (or parts) from the weights, the cameras "scan" the part and determine the size, number and nomenclature of the MTP by comparing data obtained using computer vision with images and weight and size characteristics of typical MTP previously loaded into the WMS system.

When measuring several parts in a group package on a pallet, the cameras automatically determine the presence of a pallet and subtract the weight of the pallet from the total weight of the measured parts.

Measurement reports are sent to the WMS system, and are used to account for MTP, control robotic loaders, and take into account the storage area occupied by specific MTP.

The time of mass-size measurement of one MTP or pallet is 5 seconds without taking into account the time of placing/removing the MTP on the scale.

The complex's artificial intelligence algorithms not only record an error if the cameras are not in front of the part that should be according to the information (invoice or order) from the WMS, but also determine which "wrong" part was delivered, if such was previously uploaded to the system database.

Development of prototype PAK for automatic measurement and tubing length control

By order of one of the production divisions of PJSC Gazpromneft, the company VizorLabs developed and tested a prototype of a software and hardware complex for automatic measurement and control of tubing length (tubing) directly during descent and lifting operations (SPS) when working with tubing at the well.

The solved problem

The current technology of tubing measurement at ACT requires a large amount of manual labor and significant time costs. It does not provide the necessary accuracy of measurements and metering, which leads to loss of production time and repeated repairs.

Before the tubing is lowered into the well, the assistant drill master roulette measures the pipe and records the readings in the log. After the tubing is lifted out of the well, the Assistant Roulette Drillmaster measures the length of the pipe and records the readings in a log. After completion of the work, the log with data on lowered and raised pipes is transmitted to the master in charge of the work.

• Long preparation time for each ACT: measurement of 1 tubing takes from 10 to 30 minutes.
• Incorrect determination of the length and number of tubing used by current measuring instruments: the mismatch of the actual length is +/-10%.
• Large error - up to 10 m per 1 km of tubing - in case of pipes immersion in the well during ASW. As a result - errors when entering the productive formation and the oil zone, the thickness of which is 5-10 m.
• Increase in the number of repeated works and unproductive time due to erroneous data during manual measurement of tubing.

Decision

The software and hardware measuring complex with video analytics technology performs automatic measurement of tubing pipes at ASW, counts submerged/raised pipes and determines the total length of all tubing pipes during current and overhaul of oil and gas wells. Measurement reports are generated and sent automatically to the site wizard.

The complex is located on a mobile drilling rig and consists of a video camera, a measuring complex, a target for the measuring complex, a video analytics server, Wi-Fi access point.

Implemented functionality:

• measuring the number of tubing pipes in the process of their lowering and lifting from the well;
• measuring the total length of tubing pipes in automatic mode;
• tracking the depth at which the tubing string is located at each time.

Value

Economic effect on 1 well: reduction of labor costs by 3 million rubles/year; increase of revenues from oil production increase up to 40 mln RUB/year.

• Improved measurement accuracy
• Increase of oil production volumes due to more accurate inflow of tubing into oil-bearing formations.
• Reduction of time for preparation works at RNC/winder ACS (up to 30 minutes per 1 tubing)
• Reduction of the number of repeated works, unproductive time, reduction of human impact
• Increase of oil production up to 2 thousand tons per year per 1 well

Development of PAK prototype for industrial safety control during loading and unloading operations

By order of one of the production divisions of PJSC Gazpromneft, the company VizorLabs developed and tested a prototype of a software and hardware complex for automatic measurement and control of industrial safety during loading and unloading operations (PRR) and construction work on a crane.

The solved problem

The specifics of RLP and construction works at remote sites do not allow to ensure effective control of industrial safety without significant additional costs.

• Necessity of monitoring in the personal presence of the HSE inspector;
• Inability to perform continuous and continuous monitoring;
• Analysis of incidents instead of preventing them;
• Limited possibilities for prevention of violations;
• Direct losses and damage to people's health from incidents.

Project purposes

• Continuous online monitoring of TB and BP during RBP;
• Automatic notification of TB and SPS specialists about violations via SMS and E-mail;
• Recording and storage of video fragments of cases of BOP violation, as well as data on them;
• Centralize violation information through a common database on the central server;
• Accumulation and analysis of incident information for any desired period;
• Proactive measures to prevent incidents instead of reactive response (analysis of the incident).

Decision

Software and hardware complex with video analytics technology with installation on a mobile autocrane, including:

• 2 video cameras - on the crane boom and on the crane operator's cockpit;
• Video analytics server and video analysis software;
• WiFi switch (access point) for boom camera;
• WiFi/LTE router - Ethernet switch.

The solution provides the following functionality:

• Detection of cargo, slings on cargo/hook;
• Detection of people in dynamic hazardous areas: under boom, under and around cargo;
• Instant transmission of hazard alerts to the crane operator's cabin and central control room;
• PPE wearing control (helmets and vests);
• Remote TB and BP monitoring and centralized statistics collection.