AI-powered digital platform will monitor coal supplies for the heating season

Kazakhstan is preparing for winter not only with physical fuel reserves, but also with a new tool to monitor its movement.

AI-powered digital platform will monitor coal supplies for the heating season

A unified digital platform with artificial intelligence elements should detect disruptions in coal supplies in advance before they turn into local shortages, Kapital.kz reports.

The essence in brief

  • For the 2026–2027 heating season, Kazakhstan needs to provide the population and the municipal and household sector with 7.5 million tonnes of coal — this was announced by Deputy Chairman of the Committee for State Energy Supervision and Control of the Ministry of Energy of the Republic of Kazakhstan Yernar Khasenov at a briefing at the Central Communications Service.
  • Energy enterprises' warehouses have already accumulated 4.2 million tonnes of coal and 137 thousand tonnes of fuel oil — stocks are 2% higher than last year's figure on the same date.
  • The Ministry of Energy launched the unified digital platform with AI elements and predictive analytics in August 2026 — it combines data from 36 coal-mining enterprises, transportation using 29 thousand gondola cars, 586 coal sidings, and information on more than 1.2 million consumers.
  • In the previous 2025–2026 heating season, about 8.15 million tonnes of coal were supplied to the regions — 109% of the planned volume.

What the digital platform can do

According to the Chairman of the Committee for State Energy Supervision Yerzhan Yertayev, the platform's task is to make the entire supply chain transparent and visible at every stage, and most importantly, to identify deviations in advance and take measures before they affect the provision of the population. Before the unified system appeared, information on different parts of the chain — extraction, transportation, sales — came from several disparate sources, which made it difficult to respond promptly to disruptions. Now the platform tracks the movement of coal from extraction and shipment to delivery to a specific region and sale to the population, and coal-mining enterprises are obliged not only to fulfill regional requests but also to maintain reserve capacity in case demand turns out to be higher than initial forecasts.

Stocks have already been accumulated at the start

The starting figures for this season look slightly better than last year's: 4.2 million tonnes of coal in energy enterprises' warehouses — 2% more than on the same date a year earlier. The experience of last season inspires a certain optimism — at that time, the supply plan of 8.15 million tonnes was not merely fulfilled but exceeded by 9%. According to Yertayev, the headquarters for monitoring supplies will continue working until the end of the entire 2026–2027 heating season, and not only during the initial preparation stage for winter.

Part of a broader picture of energy digitalization

The emergence of such a system logically continues the broader trend of digitalization in Kazakhstan's energy sector: in the summer, the Ministry of Energy already approved the professional standard "Digitalization and the Application of Artificial Intelligence in the Energy Sector," introducing, among other things, the specialty of engineer for Smart Grid solutions. The relevance of such monitoring systems also increased sharply after the August blackout that affected four countries of Central Asia at once — the denser the digital infrastructure of the energy industry becomes, the more important the ability to detect deviations in advance rather than react to a failure that has already occurred.

Author's conclusion

The value of such a platform will be determined not by reports on the number of connected enterprises and gondola cars, but by a concrete practical result in winter — whether it will actually be possible to prevent local coal shortages where supply chain disruptions previously were discovered only after the fact, when fuel was already failing to reach the consumer on time. For now, starting stocks look slightly better than last year's, and the previous season ended with the plan exceeded, which gives the system fairly good initial conditions for its first full-scale test in real winter conditions.