CDIO 2017 #Assignment 1

# SAE International formal classificiation system of autonomous vehicle
Figure 1 identifies the SAE formal classificiation system of autonomous vehicle, ranging from level 0 (not automated) to level 5(fully automated). Level 0 indicates that driver still takes complete control over all driving task even with assitance of warning or intervention system. In driver assistance level, the driver assistance system could either help drivers with steering or acceleration or braking determined by the environment information however still expect driver mainly responsible for all driving task still. Level 2 with partial automation, the vehicle could execute driving tasks with multiple assistance systems, however driver still be mainly in charge. From level 3, the system monitors environment information and execute autonomously with supervision of drivers but with different extent of driver involvement. As for level 3 conditional automation, human driver still need to respond propriately when system requests drivers to invervene and take controls under the conditions that could not be handled individually by vehicle. While level 4 with high automation of vehicle, driver is less reponsive to the critical situation that system could deal with that even if driver does not respond correctly. On the highest full automation level, the vehicle performs all the driving tasks full time and all conditions, therefore drivers no longer need to execute any driving tasks or monitoring environment.
Figure 1. SAE formal classification system of autonomous vehicle

# Examples of autonomous vehicle
Google's self driving car project is one of the earliest autonomous vehicle project and maybe the most successful one. The project not only cover one brand of vehicles, but different types of cars including Toyota Prius, Audi TT, and Lexus RX450h and also they has developed their own customerized vehicles. Google's self driving vehicle could execute most driving tasks in most city traffic with supervision of  safety drivers on boad in case of critical situation entire time. Googles has already performed several round of test on the real road and test tracks. According to recording, google reported a total of 636,969 miles in 2016 alone and most tests are done through city traffic in California and Texas.  During the test, the prototypes' speed will not exceed 40 km/h and driver is mainly responsible when system fails to operate safely. Therefore, based on the definition for different level of autonomous vehicle it could be categoried in level 3 conditional automation.
The second example is volvo XC90 which equip with the "Drive me pilot" system. On ordinary 90 series vehicles, they have already equiped with semi-autonomous functionality called Pilot Assist, which gives steering input to keep car properly aligned within lane markings. The new Drive Me vehicles will add hands-off and feet-off capacity in special autonomous drive zones powered by "Autonomous Driving Brain". Within the special drive zones, vehicle is responsible for the most of driving tasks but sometimes need human to take over controls if system detects unexpected situations or fail to give appropriate response which could be categoried in 4th level of autonomous system. However, out of that special regions, the vehicle could partially operate itself by detecting the environment but needs human to handle the situations that system might have some risk to execute wrongly. Therefore, it could be categoried into 2 level that is partial automation. Volvo currently launch 100 test vehicles in the gothenburg city to test auto pilot system in real city traffic. The data published now is still rare and it is hard to evaluate or determine whether the system works well or not.

# Possible application for autonomous vehicles
One of complex driving tasks is parking. With autonomous vehicle, drivers no longer waste a lot of time to find the parking space and even to have small accident due to careless execution. Meanwhile, autonomous vehicle definitely will save a lot space not just parking a lot but also space on normal road since vehicles could keep less distance between each others.
The new mobility models definitely will be brought by autonomous vehicle especially in car sharing mode, i.e taxi alternatives or car rentals. Nowadays, Uber cooperates with Volvo to study how chaufferless vehicle to take passenger to their destination therefore decreases human cost and improve system operation efficiency. For car renters, they have similar problem that it costs a lot to transport vehicles place to place. However, with the autonomous vehicles, no more hiring is needed to take care about the vehicle transportation, and the renting could be arranged in more efficient ways.
Except from alternation of mobility, autonomous vehicles changes the car-service landscape and offers more opportunity of bussiness for car OEMs. Without waste much time on waiting for maintainance or services, owners of vehicle would more likely to send vehicle directly to OEM service processes instead of independent ones. Besides, autonoumous vehicle could not only service itself but also performs other possible tasks i.e. fetch the delivery to save drivers' time.



Comments

  1. Thanks for your work on this assignment. - The CDIO Academy Team

    ReplyDelete

Post a Comment

Popular Posts