Autonomous Vehicles: From Cool Tech to Societal Transformation and Service-Oriented Deployment

Emilio Frazzoli, CTO of nuTonomy, outlines his vision for autonomous vehicles, moving beyond the initial "cool technology" phase to a focus on profound societal impact. He traces his journey from working on UAVs and DARPA Grand Challenges to realizing the transformative potential of self-driving cars through a project on urban mobility in Singapore. Initially, common justifications for autonomous vehicles centered on safety, convenience, improved access, efficiency, and environmental benefits. However, Frazzoli argues that while these are valid, they primarily represent an incremental improvement to the existing mobility paradigm rather than a fundamental shift. He emphasizes the need to leverage this technology to truly change how we conceive and utilize mobility.

Frazzoli presents a compelling economic analysis, comparing the societal value of increased safety (estimated at nearly $1 trillion annually in the US) with the value of time regained from not driving (estimated at $1.2 trillion annually). He highlights that the economic benefits of a truly effective car-sharing system, enabled by autonomous technology, could add another significant chunk, estimating a "sharing factor" where one shared vehicle replaces four privately owned ones. This re-evaluation suggests that the greatest value lies not just in making current driving safer or more convenient, but in unlocking new models of mobility. He critically assesses the Society of Automotive Engineers (SAE) levels of automation, particularly levels 2 and 3, where human supervision is still required. He argues that these levels are inherently problematic, going against human nature and potentially leading to new failure modes, drawing parallels to the airline industry's experience with autopilots and the subsequent loss of pilot situation awareness.

Instead, Frazzoli advocates for a direct leap to Level 4 or 5 full automation, where the vehicle operates without human intervention. He contends that this full automation is essential to truly capture the technology's value, as it enables the "game-changing feature" of cars moving without anyone inside. He distinguishes between two primary paths for autonomous vehicle deployment: the OEM path, which incrementally adds features from Level 0 to 5, and the nuTonomy/Waymo path, which focuses on full automation from the outset, starting with geofenced applications and scaling up. He strongly prefers the latter, arguing that the OEM path's reliance on human supervision at intermediate levels is a dangerous "red band."

A crucial distinction Frazzoli makes is between autonomous vehicles as a consumer product versus a service. As a consumer product, the cost of the autonomy package must be low, and the vehicle must operate everywhere, requiring extensive and costly global HD maps and consumer-level maintenance. As a service, however, the financial comparison shifts from the car's cost to the cost of a human driver (around $100,000 per year), allowing for more sophisticated and expensive technology. Furthermore, a service can operate within defined geofenced areas and specific conditions, simplifying mapping and maintenance, which can be handled by professional crews and even self-maintained by the fleet's data collection. He concludes by predicting that autonomous vehicles as a service will be common within a couple of years, while as a ubiquitous consumer product, they are still two decades away, emphasizing the profound impact on mobility supply rather than just job displacement.

I started thinking more about why do we want to have self-driving vehicles.

What these people do not realize is that most of those people are actually you know fairly young like in their 20s and 30s.

These are all good reasons but these are all ways that you take the status quo you know how cars are used today and you make it a little bit better maybe a lot better but you do not make it different.

The value to society of getting the time back from having to drive is actually more than the value to society or increased safety.

Car sharing or you know vehicle share in general is a concept that everybody loves but nobody uses.

Level 2 and level 3 that is anything where you require the human to pay attention and supervise the automation and be ready to intervene with no notice or with some ambiguously defined you know like a sufficiently notice they just go behind you know go against human nature.

The one game-changing feature of these cars is the fact that these cars now can move around with nobody inside.

There is a big difference between autonomous vehicles self-driving cars is a consumer product versus a service that you provide you know to two passengers.

HD maps my expectation is that HD maps within a few years will be a dime a dozen.

Mobility services worldwide are actually meant power-limited.