For the first time in world practice, Russian specialists have managed to create a technology that uses artificial intelligence to optimize the planning of public transport operations. This is not just a laboratory experiment, but a fully functioning system tested on Moscow routes.
The essence of the technology and initial results
The development, named the "Opturan" platform, was created by the Federal State Unitary Enterprise "ZashchitaInfoTrans," which is under the Ministry of Transport. The key innovation is that algorithms have for the first time managed to combine three critically important factors into a single model: the schedule of operations, the deployment of vehicles to routes, and driver shift schedules. Previously, this task was considered practically impossible due to the high complexity of the calculations.
The experiment was conducted over two weeks on nine routes of the "Andropova" bus depot in Moscow, with the participation of the State Public Institution "Transport Organizer." The results are impressive: inter-route idle time was reduced by 34%, and the nighttime workload on drivers significantly decreased. Moreover, the optimization freed up two buses, which were redirected to strengthen other, busier routes.
Strategic direction: AI in all sectors
This success is not a coincidence, but part of a systematic state policy. President Vladimir Putin had previously instructed the more active implementation of AI for forecasting demand and optimizing transport schedules. Now this instruction has received concrete practical implementation. Transport is becoming another sector where algorithms take on calculations that are beyond human capability in terms of data processing speed.
Similar processes are underway in other sectors. The Ministry of Construction, for example, plans to transfer half of the processes in construction and housing and utilities to AI by 2030, moving toward unified digital platforms. In parallel, the Ministry of Digital Development and the Ministry of Industry and Trade are developing standards for integrating AI agents into industrial software to avoid creating closed ecosystems.
The authorities are also working on building their own technological base: requirements for localizing AI chips are being discussed to reduce dependence on imports. Without domestic computing power, scaling projects like the Moscow transport model across the entire country will be extremely difficult.
Neighboring countries are also keeping pace. In Kazakhstan, two ministries have launched an AI service that independently analyzes legal norms, reducing the burden on businesses.
My view: The success of the Moscow experiment is an important signal for the entire market. The 34% efficiency is not just a number, but proof that AI can solve tasks previously considered unsolvable. However, the key challenge remains infrastructure: scaling such solutions across the entire country will require significant investments in domestic computing power. Without this, we risk getting isolated successes instead of systemic transformation.