Belarus develops a neural network that helps compile public transport schedules
A neural network has been developed in Belarus that helps compile public transport schedules. Specialists from the United Institute of Informatics Problems of the National Academy of Sciences of Belarus are working on the technology, "Minsk-Novosti" agency writes.
The system takes into account a large number of factors affecting transport operation: passenger numbers at different times, traffic jams, weather, seasonal changes, mass events, and other circumstances. Based on this data, mathematical models and the neural network can select the most optimal schedule option in a few minutes.
As Dr. Mikhail Kavalyou, Doctor of Physical and Mathematical Sciences, noted, modern cities are becoming increasingly complex for planning transport routes, as the number of factors that need to be considered is constantly growing. The new system is capable of working with tasks involving thousands of variables and significantly reduces calculation time.
However, the neural network should not replace dispatchers. Specialists will control the options proposed by the system and, if necessary, make adjustments to them, taking into account the real situation on the roads.
In the future, the technology can help make bus and trolleybus movements more stable, reduce transport waiting times at stops, and more efficiently distribute vehicles along routes. It can be especially useful in non-standard situations — for example, during large events or with sudden changes in road conditions.
Thus, the neural network should become an additional tool for transport specialists, allowing them to find optimal solutions faster. But the final decision on the schedule still rests with humans.
Now reading
In the Mironets family, the mother and three sisters have already been convicted for political reasons. Now the husband of one of the sisters has been arrested 44
Comments