Hands Free Automation for the Masses – But What is the Value of SAE Level 2 +?

In February 2020, GM announced that its Super Cruise semis-autonomous system would graduate from the Cadillac brand, eventually becoming available on 22 GM vehicles by 2023. Precise details on which models will feature Super Cruise have yet to emerge, but GM has signaled that its larger, higher-volume models will feature the SAE Level 2 highway automation system, which currently costs Cadillac customers an additional US$2,500. GM, therefore, has become the first Original Equipment Manufacturer (OEM) to transplant a semi-autonomous system from its low-volume premium brand into higher volume mass-market brands.

In the space of 10 years, the mass market vehicle segment will therefore transition from passive safety to almost universal adoption of basic active safety collision avoidance and, finally, to the green shoots of semi-autonomous driving. Within the broader strategy of democratizing SAE Level 2+, there are a number of avenues open to OEMs.

It was always expected that OEMs would eventually look for a broader cost distribution for their autonomous vehicle systems, which have been engineered at great expense over the course of the last 5 to 10 years. While many OEMs had intended to monetize their autonomous vehicle systems via mobility services enabled by driverless vehicles, the realities of technology limitations, regulatory lag, and the rapid rise and collapse of OEM-lead shared mobility initiatives are pushing OEMs in the direction of Level 2+. In practice, a vast spectrum of longitudinal and lateral assistance can be yoked under the term Level 2 which, happily for OEMs, keeps the driver responsible and liable, but also restricts the additional value that OEMs can deliver to the consumer.

In deploying Level 2+, and in further developing their autonomous systems within the scope of Level 2+, OEMs can commit their limited resources to any of the following strategies:

Enable Hands Off via DMS: In all L2 applications, the human driver is a Safety Critical Element (SCE). Most L2 implementations have policed the human element though crude torque sensors on the steering wheel. By opting for a camera-based Driver Monitoring System (DMS), GM can ensure that the driver is cognitively engaged, allowing the driver to remove their hands from the steering wheel. While “hands off” is a vital stepping stone on the path to L4, requiring the driver to have “eyes on” and “brain on” dilutes the value for the consumer.

Increase Operational Domain: L2 systems were first introduced to market as the combined functionality of two distinct Advanced Driver-Assistance Systems (ADAS): Adaptive Cruise Control (ACC) and Lane Keeping Assistance (LKA). The operating conditions of such systems were very limited, with ACC often not able to support start/stop and LKA only supporting limited lane curvature. Therefore, many OEMs have focused on expanding the operational domain of L2+ to accommodate a greater variety of speeds, more severe bends in the roads, and a larger number of road types. For example, GM recently extended the operational domain from 130,000 km to 200,000 km of United States roads via the expanded coverage of its HD mapping partner Ushr. In such implementations L2+ is differentiated form L2 through greater reliability and availability, adding value for the consumer.

Enable Automatic Lane Change/Highway Exit: In addition to holding the vehicle within the lane of travel, L2 systems can also execute maneuvers on behalf of the driver, such as overtaking, lane changing, and taking exit ramps. In Tesla’s current implementation, this still requires some input from the driver, such as moving the indicator stalk.

The three above strategies represent a departure from the decades-old status quo, and while all three can be independently implemented within a L2+ framework, it will take a combination of all three to advance into higher levels of automation. Therefore, while “hands free/brain on” investment may deliver limited returns, the development of robust DMS will be key to advancing from L2+ to L3. OEMs must always have one eye on the challenge of advancing beyond L2 to L3 or L4, as the business case for selling consumers an L2+ system, which has little functionality delta with L2 but a considerable cost delta, is far from clear. The strategic reasoning behind the L2+ trend is the desire among OEMs to begin monetizing their autonomous stack in the short term, rather than waiting for regulatory clarity or validation of complex Artificial Intelligence (AI)/algorithms.

Given that the L2+ stack, from both a hardware and software perspective, is overpowered for basic L2 applications, OEMs should pursue a strategy of leveraging this “headroom” to deliver functional improvements Over-the-Air (OTA), potentially even enabling a jump from L2+ to L3, thus providing a potential aftersales revenue stream. While expanding the operating domain of L2+ provides more immediate term value for the consumer, equipping models with DMS and 360-degree perception provides scope for safe upgrading to eyes-off automation. In many ways this flies in the face of conventional automotive thinking, whereby the very minimum hardware specification to enable the desired functionality at point of sale is invariably selected to minimize costs. However, as cybersecurity considerations require life-cycle management and force a relationship between OEMs and their models post-sale, OEMs must come to see OTA as a revenue opportunity, and not just a remote software maintenance channel.