Perspective: Self-driving cars as an ideal is not the same thing in real life

1
self driving car

At the end of May, Google created a flap with a media blitz displaying its latest iteration of self-driving cars. While the “bubble car clearly engendered a host of interesting discussions, the Google people suggested that, although their technology could produce acceptable results in one or more controlled environments, it would be some time before the product would be ready for prime time.

Given this fact, if the same guys who spent years, countless R&D millions, and thousands of billable hours trying to create consensus to sell the idea of autonomous vehicles to the public and State/Federal regulators are timid about their production plans, one might also think that companies who are following Google’s lead would follow suit and, therefore, be careful with their own rollout programs. However, in the world of digital transportation technology, technical capability and practical prudence are many times mutually exclusive; particularly when it comes to small, highly-innovative companies who are sometimes inexperienced with practical concerns driven by the potential failure of their products.

In the end of the day, then, this leads me to a new bolt-on auto-drive system called Cruise.

According to the company, Cruise is “a highway autopilot” that “controls your steering, throttle, and braking, making sure your car remains safely in its lane and a safe distance from the car in front of you.”

The product appears to be very early-to-market when compared with the rest of the auto-drive development segment, and consequently, the product could end up creating a first 21st Century illustration of Murphy’s Law where “anything that can go wrong will go wrong.” More problematically, however, if ‘whatever can happen’, ‘does happen’ the overall justification for autonomous vehicle development could be damaged for years or worse.

In order to factor just one major concern associated with the premature sale of Cruise, consider the legal liabilities, along with the negative perceptual impacts created by the system going wrong. Today, any bit of information, rumor or otherwise, hits Twitter, Pinterest, Instagram and YouTube in a heartbeat. These opinion-driven information outlets nevertheless produce “factoids”, governed only by the legitimacy of 140 character word packaging, or imagery that may, or may not be contextually accurate.

Unfortunately, however, once these “factoids” hit the Net, they stay on the Net – forever. This means that perceptual differences between “accurate”, and “entirely false” are usually the width of a hair; so getting a company’s products done right the first time it highly critical, lest an enterprise’s lifespan resolves itself upside down financially, in court, or all the above. On top of these concerns, Cruise bolts on to existing vehicles, and practical experiences with newly released aftermarket auto products are typically shown to be problematic; particularly when they relate to early-stage technologies.

cruise
Cruise Automotive

Now, all that said, in order to establish a Cruise failure scenario, let’s assume for a moment that a customer’s system suddenly goes into fault mode while entering the San Mateo-Hayward Bridge. This facility is known to be the second major crossing point between populous San Mateo County and the San Francisco Peninsula, after the Golden Gate Bridge.

The speed limit along that corridor is 65 MPH, and in our story, the Cruise car’s owner is a wealthy stockbroker who is much more interested in riding un-attended while executing real-time trades on his iPad, rather than paying attention the fact that the car’s auto-drive system just dropped offline, and is now drifting out of its lane. At the same time, let’s assume that the car is likely to be the only auto-drive system on the bridge, running in close proximity with hundreds of other human-controlled cars who are also running down the bridge at 65 MPH. Consequently, then, what do you think would happen during, and after the expected crash?

My guess is that one or more cars would be upside or in the water, before the stockbroker could even respond to the initial threat physically. Then, assuming that he was not killed outright, he or his family would have the unenviable responsibility of responding to long lists of costly lawsuits which would, in turn, involve the system’s primary manufacturer, along with other involved supply-chain manufacturers, drivers and passengers, the Bridge itself, San Mateo County, and potentially the State of California, NHTSA, and even the Federal government at-large.

Consequently, the point of this allegory is simple and straightforward; with today’s technology virtually everything is possible. However, just because a thing “can” be done, it doesn’t necessarily follow that it “should” be done; particularly until a technology is entirely proven in the real-world.

That’s why in the case of auto-drive, Google is continuing to operate slowly, and on the basis of a series of well-controlled practical programs; where, on the other hand, Cruise’s premature public rollout is likely to turn their early buyers into test dummies, rather than satisfied customers. Granted it’s all about risk-management, but if I had to put my family in an auto-drive vehicle developed by a company that harbors a market cap well beyond $300 Billion, versus eight guys in a business park in San Francisco, I’d take the big company any day.