Apple's preliminary data on retreating its self-driving car project had surfaced yesterday, pointing to a large number of retreats. Today, the DMV has shared the company's complete retirement reports and provided insight into Apple's Autonomous Autotests.

A release report keeps track of the number of times an autonomous vehicle is disengaged and restores control to a safety driver or how often the safety driver intervenes in the vehicle.


Yesterday's information indicated that Apple had the worst ranking when it came to retreat, but Apple has now specified details [PDF] Explains the reporting procedures for the withdrawal and some changes made in the middle of the year.

For the period between April 2017 and June 2018, Apple vehicles drove 24,604 miles on their own and experienced 40,198 manual takeovers and 36,359 software deactivations, a figure that is comparatively high based on feedback reports from other companies.

However, in July 2018, Apple stopped listening to the total number of unsubscriptions and instead focused on "key deductions," also known as "retreats," which may be a safety incident ("accident") or violation of road traffic regulations have led .

With this metric, Apple vehicles have driven 56,135 miles since July 2018, with only 28 "Important Disengagements" reported. Two of these "important shutdowns" were actually smaller collisions that were not to blame for the vehicles of Apple. One accident occurred in August 2018 and the other in October 2018.

At Apple's revised reporting threshold, the company's cars experienced only one major retreat every 2005 miles, compared to every 1.1 miles when all the data was counted. If other companies use similar thresholds to Apple's new standard, Apple would be in a much better position.

It is difficult to make direct comparisons between the Apple Takeover Report and the results of other companies, as there is no standard for reporting on withdrawal. It is up to each individual company to decide what a deduction means and what deductions have to be reported.

However, it is clear that Apple's vehicles are still at an early stage of testing, as the company itself says in a DMV cover letter.

According to Apple, security is "the highest priority", and the retraction mechanism is "conservative" because the system can not be used in all situations and situations.

Apple's test parameters assume that the driver takes on manual control of a vehicle each time the system encounters a scenario that goes beyond the current capabilities. The vehicle itself monitors itself and returns control to the driver if any errors or problems occur.

Situations in which the driver takes control include the appearance of ambulances, construction zones, or unexpected objects on the road, as Apple vehicles can not navigate these obstacles themselves.

The autonomous software provides control when it can not track an object sufficiently, can not create a motion plan using the path planning system, when the vehicle systems are not responding as expected, and communication problems occur.

Apple now has more than 62 vehicles on the road, a number that is expected to increase in 2019 with the autonomous testing of software. Apple has to hand over annual retreat reports to the DMV so that by the beginning of 2020 we will see the company's performance for 2019 and look for improvements.