Simpl has developed a logistics optimization solution for gas supply in the oil and gas industry. Using machine learning tools, the system has helped cut supply planning time fivefold and reduce transportation costs.
Large oil and gas companies operate multiple gas production sites and serve vast numbers of consumers. Gas can be delivered to customers through different routes: by land, pipeline, or rail. Each option comes with its own transportation specifics, costs, and infrastructure limitations across regions. For example, some pipelines have limited throughput capacity and cannot deliver the required gas volume at a given time.
Previously, logistics teams had to manually create calculations, reports, and summary tables for each option, entering all data into Excel by hand. They compared transportation routes one by one, cross-checked figures, calculated the economic benefit of each scenario, re-entered data, and recalculated multiple times to avoid mistakes. The analysis was based on available logistics capacity at the moment of shipment. At the same time, routes had to be designed in a way that remained economically viable.
The cost difference between transportation options could amount to several million rubles. When evaluating the labor costs for developing an automated solution, the client found that such a system would be cheaper than the potential losses from incorrect transportation choices.
The challenge was to automate the process of transporting raw material from point A to point B, while analyzing all possible routes and identifying the most cost-effective solutions.
Simpl developed a software platform where users input the production volumes from gas extraction sites and the required amount of raw material for customers. Using ML algorithms and a mathematical optimization model, the system selects the optimal and most economical delivery option in seconds. It automatically distributes gas among delivery points and analyzes data such as:
-
supplier and customer pricing,
-
raw material costs,
-
transportation costs,
-
volume allocation and profitability for supply planning.
Logisticians no longer need to calculate transport economics manually—they simply enter production and demand figures into the system. Without the solution, staff could evaluate only 3 transportation scenarios in 4 hours; with the model, they can process 1,000 scenarios in just 5 minutes. The system independently identifies the best option and generates a final report.
