Our company won the Gazprom Neft Innovation Challenge 2018 contest of innovation projects on IT topics relevant for Russian oil and gas companies. The contest is organized by the Skolkovo Foundation.
We have successfully completed the first stage of the Digital Drilling project on introduction of modern smart technologies for automation of decision-making processes for well drilling. The project is jointly implemented by Gazprom Neft-Yamal (Tyumen) together with ITSK (St. Petersburg) and Smart Solutions (Samara).
The aim of the project is to develop solutions that allow users to analyze the current situation, taking into account objective information, which is being formed in real time, and then allocate, plan and optimize resources, monitor and control the progress of well construction, adaptively react to deviations from the schedule and extract knowledge from experience.
Our company created a prototype of the knowledge base and multi-agent system for building and adjusting the drilling plan in real time based on incoming data, accumulating and taking into account best practices of drilling teams operating under similar conditions.
Multi-agent technology provides the ability to flexibly and efficiently form multi-criteria process schedules, finding the balance among conflicting interests, preferences and limitations of agents of orders, resources, tasks of technological processes, workers, equipment and materials.
The most important domain data, which determines the decision-making logic of agents, is placed in the Knowledge Base and separated from the source code. Thus, this information can be replenished by field specialists.
Examples of knowledge base and system schedule
Our multi-agent system automatically builds the plan for drilling operations, taking into account all accumulated changes in technologies and adaptively reacts to unforeseen events in order to avoid delays. The system flexibly rebuilds plans in response to any events, which makes them always relevant. The system also automatically builds planned indicators of the use of working time, machine loading, and operation execution, allowing users to analyze the well drilling processes as the situation progresses.
The system prototype has been successfully tested on examples of real well construction plans and made it possible to automatically build work schedules that reduce time costs due to more efficient use of resources and best practices in carrying out technological operations.
Gazprom Neft-Yamal plans to implement our adaptive scheduling system for well drilling in industrial operation this year, which will help the company make an important step from automation of physical drilling processes within Industry 4.0 to digitizing domain knowledge and automating decision-making processes within Industry 5.0. It will also reduce the project implementation time and resource costs, and provide the opportunity to fully use best practices in construction of wells.
Moreover, the developed system has helped Knowledge Genesis Group of Companies in becoming a resident of the Skolkovo Technopark, for which re-registration is currently in progress.