Published on: 2026-06-05
Source: Saint Petersburg Polytechnic University of Peter the Great –
An important disclaimer is at the bottom of this article.
We continue to talk about the team of the “Priority 2030” program. Today we are discussing the project “Automation of seismic data processing using artificial neural networks” under the leadership of the chief engineer of the laboratory “Digital modeling of underground oil and gas reservoirs and well-test analysis,” Ivan Zhdanov.
Artificial intelligence in oil and gas geological exploration is not a new task, but for a long time it was solved fragmentarily: separate algorithms for interpolation, separate ones for noise suppression. The head of the project “Automation of seismic data processing using AI” Ivan Zhdanov decided to combine these fragmented solutions into a single system. This gave birth to a project for automating seismic data processing and predicting reservoir properties in the interwell space.
The team was assembled out of interest: there were programmers, geophysicists, and mathematicians. Many came from the master’s program at the Research and Education Center “Gazpromneft-Politech.” Dmitry Pashkovsky, for example, started by writing code for the first neural network models, and Igor Gruzdev worked on data processing. Gradually, a core was formed that could solve the tasks “from idea to code.” Currently, the Research and Education Center employs more than 90 engineers who possess deep industry expertise. It is precisely they who make up the core project teams.
According to Ivan Zhdanov, the main thing is to see how algorithms that only worked on synthetic data a year ago are now successfully passing approval with an industrial partner.
“We create programs for computers that truly accelerate the processing of geophysical data from weeks to minutes,” comments Dmitry Pashkovsky.
The funding of the “Priority 2030” program allowed the team not only to acquire computing power but also to focus on long-term tasks: additional training of neural networks on real data, integration of three modules (GNNTransformerPetro, GenPetroConverter, JointFWI) into a single platform “Polanis.” Thanks to “Priority 2030,” the team completed three software registrations for computers and published several scientific articles, including publications in the “White List” journal.
Our goal is to obtain an integrated solution by 2027 that the customer can simply take and integrate into their production conveyor. “Priority 2030″ gave us that very long horizon, which is usually lacking in applied research,” noted project manager Ivan Zhdanov.
Please note; this information is unprocessed content obtained directly from the information source. It represents an accurate report of what the source claims and does not necessarily reflect the position of MIL-OSI or its clients.