Toyota Says Automated Driving Grading Causes Confusion | Urgent Need for Clear Level Differences

According to an AutoCar website report, the person in charge of the Toyota Research Institute said in an interview with reporters recently that due to the current progress in the development of autonomous vehicle systems, the current classification system for self-driving cars has caused confusion.

The current grading framework was originally proposed by the SAE International in 2014. The framework divides the automatic driving level into six levels, which are 0th to 5th. The 0th level autopilot system can only be When the vehicle encounters an obstacle, the driver is alerted, and the fifth-level automatic driving system can completely autonomously drive by the vehicle without manual intervention.

Most current car manufacturers are studying Level 2 (supporting the driver's hands out of direction), Level 3 (supporting the driver's line of sight from the driving line) and Level 4 (supporting the driver's attention away from the driving motion). System. The director of the Toyota Research Institute, Jill Pratt, said that the small differences between the above three levels make the developers very confused.

According to SAE International's rating framework, the Level 2 automatic system can perform certain driving tasks, but the driver needs to monitor the driving environment and prepare to take over at any time. At present, most car companies have already achieved Level 2 autopilot technology. At level 3, the driver will no longer need hands and feet to stand by, and the machine can perform almost all driving operations independently, but the driver still needs to maintain his or her focus so as to cope with situations where artificial intelligence may not be able to respond.

The 4th and 5th level of autopilot technology can be called fully automated driving technology. At this level, the car can perform all driving operations without the need for driver intervention. The driver can also pay attention Focus on other aspects such as work or rest. However, the difference between the two is that Level 4 autopilot applies to some scenarios, usually in cities or highways. Level 5 requires auto Driving a car can be fully driven in any scene.

In an interview with reporters, he said: 'Understanding the differences between different levels of autopilot technology is extremely important. For example, level 4 autopilot technology depends on the driver's environment, the scene and the interaction between the vehicle and the sensor. '

He also stated that if a vehicle that has achieved Level 4 autopilot function has been under manual safety monitoring during the test, then the autopilot performance of this vehicle has not reached Level 4, at most it can only be regarded as Level 2. For auto-pilot technology, the core problem is to see how the technology itself can function in a variety of environments and scenarios. The existing grading places too much emphasis on the driver's involvement.

It is pointed out that Gil Pratt's remarks also reflect the driver's perception of the current level of autopilot confusion. Tesla had previously introduced Level 2 autopilot on the ModelS sedan, but it has since occurred. Due to the failure to focus on the road leading to the timely failure of the vehicle to take over and then happen.

Road test at ACM

The American Center for Mobility (ACM) recently officially opened in Westgate, Detroit. This 335-hectare center has a dedicated automated driving test facility.

Many companies have already conducted autopilot drive tests at ACM. Toyota calls itself the first car company to conduct drive tests at the center. The center will be open to many companies, which means that global car companies, suppliers and technology companies will Here to learn skills.

Inside the facility there is a 2.5-mile loop road, a 700-foot curved tunnel, two double-deck bridges, an intersection, a roundabout at the road, etc. Fiber-optic network connections are implemented on the test road, Test data can be collected to enable communication between cars, communication between vehicles and traffic lights and other infrastructure. ACM uses a complete wireless network to provide controlled test conditions.

Self-driving vehicles are the future development direction of the automotive industry. ACM will help companies develop automated driving technology, develop grading standards for automatic driving technology, test vehicle safety, and enable the public to finally accept this technology.

2016 GoodChinaBrand | ICP: 12011751 | China Exports