Just as California has called the regulatory tune as far as vehicle emissions are concerned, with the impact of the California Air Resources Board's decisions being felt around the world, the California Department of Motor Vehicles is making its voice heard regarding autonomous vehicle regulations with equivalent impact. This week, California let it be known that driverless cars can now really be driverless, as long as a few rules are adhered to.

Those rules include, according to a report in the Washington Post:

1. Certification that the subject vehicle has "been tested under controlled conditions that simulate, as closely as practiable" the types of conditions and circumstances they are designed for such as operating at night, in the rain or in a particularly busy geography;
2. Certification that the suject cars are "designed to detect and respond to roadway situations in compliance with California Vehicle Code provisions and local regulation applicable to the operation of motor vehicles.;"
3. Notification of local communities of relevant testing plans;
4. Presence of a "two-way communication link;"
5. Notification of the state in the event of a fender bender or more serious crash and/or when the vehicle's autonomouse technology "disengages" and needs, for example, to pull over because of a malfunction.

The significance of these rules lies in their ability to influence testing protocols throughout the U.S. including efforts at the Federal level, and the world. As for the specifics of the rules, the codification of actual driverlessness as acceptable is the game changer. California is finally on board with jettisoning the so-called "safety driver."

It is most notable that the new rules arrive in the aftermath of the fatal autonomous Uber crash in Tempe, Ariz. The message from California clearly suggests that Uber's failure will not impede the long-term progress of autonomous vehicle development. So far, only one state, Minnesota, has seen the introduction of legislation to indefinitely ban autonomous vehicle testing. (https://tinyurl.com/yd8ge8mz - "Self-driving Car Ban Bill introduced in Minnesota Senate")

Equally important is the inclusion in the California rules of a stated requirement for a two-way communication link. This marks the first time that connectivity is specified for autonomous vehicle testing by a regulatory body.

This finally settles a debate which has called into question the need for a connection to support autonomy. The earliest self-driving cars from companies such as Alphabet's Waymo division did not rely on wireless connections to operate. These vehicles were entirely reliant on the fusion of inputs from Lidar, radar, camera and other sensors.

What has emerged from the experience of observing autonomous vehicles operating in the wild is the need for remote control and/or remote shut down, in the event of a failure. Nvidia demonstrated at its GTC 2018 event last week in San Jose how virtual reality technology might be used to remotely control an autonomous vehicle. (picture above)

California was already the first state to require disengagement reports and reports related to collisions and the new driverless rules extend that approach. States such as Arizona have sought to create the most liberal regulatory environment possible to foster autonomous vehicle development without any data reporting requirements. The U.S. Congress, too, has passed legislation, still lacking a Senate endorsement, which skips the need for data reporting under any circumstances.

The likely outcome here is that California's middle-of-road approach - requiring reporting, but allowing true driverless operation - will prevail. Once again, California takes the lead in defining the regulatory for future transportation priorities. Regulators around the world will take note.