news

한어Русский языкEnglishFrançaisIndonesianSanskrit日本語DeutschPortuguêsΕλληνικάespañolItalianoSuomalainenLatina

it house reported on september 29 that the official weibo account of autonomous driving company momenta announced tonight that during the global smart car industry conference, momenta released its first large-scale mass-produced smart driving model.integrate sensing and planning into one large model, forming an end-to-end, deep learning autonomous driving solution.

officials stated that momenta’s large-scale mass-produced intelligent driving model can effectively solve the long-tail problem of autonomous driving (it house note: usually refers to special situations or rare events that occur infrequently but with a wide variety of occurrences) and support driving in various complex road environments.accurately predict the passing intentions of vehicles or pedestrians, automatically adjust vehicle speed, flexibly change lanes, and calmly avoid obstacles.

momenta further stated that this large model can handle complex intersections or dynamic crossing scenarios with ease, and can significantly improve driving safety and traffic efficiency. even invery narrow parking space at night, parking space on dead end roadprecise parking can also be achieved in extreme scenarios. currently, this technical solution has been delivered in mass production on multiple automobile brands.

according to previous reports from it house, momenta ceo cao xudong once proposed "moore's law of intelligent driving": software experience improves every two years. 10 to 100 times;hardware bom cost (raw material cost) every two yearsreduced by half。