Mobileye's autonomous vehicles, now equipped with the True Redundancy sensing system, show in a new and unedited video driving skills similar to those of a real driver.
Mobileye, from the Intel Group, today introduced in Israel the True Redundancy sensing system that operates without human contact. This is a significant milestone towards the launch of the robo-ceremonial services planned to operate in Israel and Germany. The autonomous vehicle shown in the video simulates the skills of a human driver and imitates multi-stop driving similar to ride hailing services. "Contrary to the norms accepted in the industry, Mobileye's True Redundancy system works with two separate sensing systems that back each other up. The vehicle's natural maneuvering in incredibly complex scenarios proves the value of this approach." says Johann Jungwirth, vice president of mobility-as-a-service, Mobileye.
The video shows Mobileye's autonomous vehicle functioning as a robo-taxi service, traveling to multiple destinations and stopping at passenger pick-up and drop-off points. In the fifth episode of the series of unedited driving videos, the full capability of the True Redundancy system is shown. Mobileye's unique approach to merging the types of sensors in the autonomous vehicle (AV) leads the company's robo-taxi in navigating the complex streets of Jerusalem at night. While previous unedited videos showed the autonomous vehicle driving with only the camera system, the new video shows the autonomous vehicle in its full configuration that Mobileye plans to use in commercial deployments of the robo-taxi.
How it works: True Redundancy is Mobileye's unique approach to sensor fusion that uses two independent sensing subsystems – one with cameras only and the other combining radar and lidar – to act as backups to each other instead of complementary systems. The result is a sensing solution that should extend the average time between failures. The autonomous vehicles currently driving in Israel are the first Mobileye vehicles that combine the two systems in one vehicle, and illustrate the expected performance of the robo-taxi in action in the real world.
In an unedited 40-minute video, Mobileye's autonomous vehicle is seen performing complex driving maneuvers on city roads despite difficult night lighting conditions and complex road signs. The human driving of the autonomous vehicle seems remarkably normal, and it manages to handle challenging maneuvers without any hesitation.
The autonomous vehicle integrates with human drivers when making a left turn at an unsignalized intersection (4:04; 06:48), which seems simple but in practice is not at all, navigating safely and successfully around pedestrians crossing the road without a crosswalk (08:28,10 :42), smoothly deals with illegal maneuvers of other drivers (04:34); completes a 180-degree turn at an intersection with multiple traffic lights (25:18); navigates around vehicles blocking the lane (25:39); navigating a traffic circle with pedestrians (26:44); and performs other normal driving maneuvers.
merging sensors
The demonstration of True Redundancy on real roads helps to remove the doubt that the industry has placed on the approach adopted by Mobileye to fuse sensors. The autonomous vehicle performs the trip as a human driver would, and demonstrates almost complete readiness for the activation of the planned robo-taxi services. Building on the formidable capabilities of Mobileye's camera-based autonomous vehicle development fleet, the addition of radar-lidar to its sensor suite was the latest step in achieving the goals the company has set for itself with its unique autonomous vehicle technology.
Making the True Redundancy system active is a major milestone in Mobileye's robo-certific service planned for deployment later this year in Germany and Israel. Mobileye has begun the process of obtaining the permits and regulatory approvals in both countries, which will allow the company to begin waiving safety drivers on public roads.
Mobileye Drive, Mobileye's self-driving system, incorporates Mobileye's industry-leading technologies, including Road Experience Management, the company's proprietary approach to mapping, which leverages crowdsourced data from advanced mass market driver assistance systems to build maps for an autonomous vehicle on short notice; Driving policy Responsibility-Sensitive Safety (RSS) which applies a mathematical model to improve safety through improved adaptation to unique driving environments; and True Redundancy, which combines two independent sensing subsystems powered by radar-lidar cameras, each capable of developing full models and ultimately supporting full end-to-end autonomous capabilities.
More of the topic in Hayadan:
3 תגובות
9:30 a.m. Traffic offense by Volt messenger. Who is preparing an application that will send online reports to these idiots?
The vision of autonomous cars will change the face of our society
A private car will no longer be required, because it will be easier to order a robot taxi, which will be available to take you from your door.
In combination with the technology of the ree company, where there is no body of a steering wheel and axles and an engine in front, but everything is in the wheels, the future looks rosy. Autonomous electric cars, with more room for passengers to share comfortably.
Reduction of traffic congestion and accidents. Improving work productivity, because it will still be possible to work, from the moment you get into the car until you arrive at work.
However, I wonder what will happen to all the driving instructors? To all public transport drivers, truck drivers and delivery drivers?
What about off-road drivers, the whole experience is self-driving and not autonomous. Will they continue to develop vehicles with hybrid systems?
Absolutely impressive
Only about 15 years ago it was hard to believe that such a level could be reached, there is definitely still a serious long way to go until we reach a complete system that can be implemented on the road,
I wonder what the path will lead to a completely autonomous system, the path of full mapping before the trip
In which as far as I know Mobileye and most companies use or Tesla's second way which is more autonomous they identify the road according to a set of road traffic laws that the system knows and identifies in the road you enter which is more similar to human driving,
There is also an overlap between the two systems and that is dynamic movement detection and reference to it seems to me to be expected in the videos
There is a closeness between the two companies, but it is difficult to compare. There are many more videos of non-Tesla people who test Tesla's driving and challenge it from Mobileye, where the videos are under the control of the company.
If the path of full mapping reaches the finish line first in terms of a robotic taxi, it may be good enough because the mapping will be carried out in the areas where there is the most financial income and only on these axes will the robotic taxi drive,
But if the 2 systems mature at the same time, it seems that there is a clear advantage to Tesla's approach
Every town, every road is a way for a Tesla vehicle to travel, all that is needed is a navigation system,
Tesla's approach could also later have an application in the robotics of robots and also where Elon Musk is trying to enter like him, everything will depend on whether they can really reach a completely autonomous system,
One of the achievements in the field of autonomous driving is that Tesla, for example, compares their vehicle to an average human driver, and as you know, human driving includes the aggressive driver and the drunk driver, etc.. But the correct comparison should be with the careful driver who has not had an accident for 50 years or more because why would such a driver want to be on A vehicle that is less safe than its driving ability, which when examined from the aspect of robotics, autonomous driving by a careful driver is certainly impressive.