[Video] Honda announces roadmap towards full autonomy by 2025 image

The Japanese automaker was among the few ones lacking concrete plans for entering the autonomous era, but it has now rectified the slip off with its new target strategy.

To be more precise, Honda is looking to achieve Level 4 automated driving capability by 2025, which is just a step behind Level 5 that has to do with completely driverless cars (as in no pedals and steering included), and on that road it will arrive with highly automated freeway driving by 2020. The technologies are already being tested at R&D center in Japan, with the important announcement being made by Honda Motor Co., Ltd. President & CEO Takahiro Hachigo – where members of the media were allowed to trial the automated vehicle technologies in complex driving scenarios, including systems with advanced artificial intelligence (AI).

So far the company established plans to reach highly-automated freeway driving capability (SAE Level 3) by 2020, and now wants to be ready with highly-automated driving capability in most driving situations (SAE Level 4) by 2025. “We will strive to achieve the technological establishment of Level 4 automated driving for personal car use by around 2025,” said Hachigo. “We are striving to provide our customers with a sense of confidence and trust by offering automated driving that will keep vehicles away from any dangerous situation and that will not make people around the vehicle feel unsafe.”

The freeway test included traffic and multiple lanes on a closed course, with the vehicle using multiple cameras, five LiDAR and five radar sensors. The second one – interestingly – simulated a common urban experience and used Honda’s latest progress on artificial intelligence (AI) with Deep Learning, there were only camera sensors and no LIDAR or GPS. Still the AI managed omplex driving environments and situations, including roads without proper lane markings, and could even work at night, detecting pedestrians and bicyclists and by accumulating experience it could improve its prediction to take better actions.