The Wheel-Steering-by-Itself Syndrome
My first encounter with a semi-autonomous vehicle was at Fremont, at the Tesla plant. My manager had just received his brand new Model 3 and we tested the Autopilot out on the parking lot. Seeing the steering wheel turned all on its own, was definitely beyond my basic human instinct.
I think today my initial reaction towards the technology in action is still shock. This is ironic. After all, I’ve written a book on self-driving cars, worked with startups developing perception stacks, and crossed paths with many Cruise pilots driving by my house in San Francisco. Yet, playing the passenger inside an autonomous machine has a tint of sci-fi experience.
In case you haven’t noticed, many of our cars today, even the less eco-friendly non-Tesla ones, already come with semi-automated features. The new sophisticated features go by the name of ADAS (advanced driver-assistance systems). These systems have progressively been introduced to enhance vehicle safety. Can you recall the subtle beeping when veering into the highway median or drifting into another lane? It’s like having a non-speaking version of your parents commenting on your driving.
ADAS features are not only becoming standard in the auto industry but are being enforced by regulators. As of 2018, NHTSA has mandated that all new cars come equipped with backup cameras. However, even with the sophisticated rearview camera, many of us still look over our shoulder while reversing. Is this out of habit or plain mistrust of the technology? I think both.
Subconsciously, having control over the steering wheel of the car gives us a sense of ownership. Everything in our lives may be chaotic but one thing we have control over is the way 3,000 lbs of steel on wheels handles as it backs out of the supermarket parking lot.
Cameras are passive devices providing the driver with data about its surroundings. It’s still up to the person to press the gas and reverse. But what happens when the sensors feed data to the central computer, which is connected to the mechanical controls of the car, and independently drives? The human is removed from the car’s maneuvering and driving escalates beyond Level 2 automation.
My mom best summarized the sentiment of the general driver in one sentence: “It’s just too weird to see the steering wheel turning on its own”.
Inconspicuously, carmakers are pushing us towards a hands-decision-free driving era. We are slowly transitioning from ADAS to its younger, smarter sibling, autonomous technology. Nonetheless, if automated rear parking makes us antsy, imagine what the reaction of a driver would be when the machine steals away the right to turn, accelerate, and halt. My guess would be fear and mistrust.
And what are self-driving companies doing to ameliorate people’s negative pre-conceptions? There is no point of maturing the vehicle technology and monetization, if there is no market for it. Surely, our roads today aren’t shy of self-driving pilots. But how many have real passengers inside of them?
Prior to the Arizona accident, Uber was already operating autonomous rides in Pittsburg, PA with users who opted for a robo-rideshare. In that city alone, the TNC (Transportation Network Company) had registered over 40,000 rides of this type. Malcom Glenn, Strategic Partnerships and Public Policy Manager, shed light on the pilot program:
Robo-taxis are the perfect use case to deploy first given current public opinion towards driverless cars. People could be apprehensive about adopting a technology with which they have never engaged with in the past. Apart from seeing X-Men’s Logan on top of an autonomous semi-truck, how many of us have come across a car that drives itself? This is why, testing with real customers can provide feedback on customer experience, perception (the human type), and behavior when driven by a robot rather than by another person.
Similarly, at the end of 2018, Google’s self-driving group started the Waymo One taxi service which charges passengers for its ridesharing service. The commercial trial, deployed in the suburbs of Phoenix, AZ, is offered only to those in their Early Rider program. Waymo uses Chrysler Pacific hybrid vans with an operator at the front wheel (to intervene in case of emergencies).
Waymo One passengers have reported that the technology didn’t seem mature enough to drive completely autonomous. They questioned the reliability of the service if the safety driver were to be removed from the driver’s seat. Despite of the driving glitches, they seemed confident about the robo-taxi’s safety and performance.
Early Riders are enthusiastic about the technology and also forgiving. But, would the recollection of the autonomous driving experience be as optimistic coming from a more neutral passenger? They’d probably be more apprehensive, skeptical and way more perplexed about the steering wheel spinning on its own.
The University of Michigan launched last year an electric autonomous shuttle on campus to observe human-machine engagement. The technology, manufactured by the French company Navya, has been tested and certified to work within a geo-fenced location. Once all the Olympic size regulatory hurdles were overcome, UM’s electric self-driving shuttle started to operate commercially to study the reaction of its patrons toward robo-vehicles. Unlike many of the OEMs’ driverless test fleets, the shuttle does provide a service to students, professors, and university staff around campus.
Greg McGuire, Director of the Smart and Autonomous Mcity at UM, emphasized that the shuttle goes beyond a technology feasibility demo:
The main objective of launching such a vehicle is not to gauge the functionality of the technology. Rather the objective of the shuttle is to observe and study how humans interact with such automated machines. Whether you are a walker, a driver, or a passenger, the emotions, reactions, and behaviors are expected to be different than the current status quo.
The main question to be answered here is, what important information can we take from your behavior that we will need to know when we start using highly automated systems with humans? I tried to put myself in the shoes of those that will either ride or encounter the e-shuttle at UM.
I think a first encounter would also differ from my second or third encounter with the smart machine. Once I stepped inside, my survival instincts would go on instantaneously. I’d want to observe the speed at which it moves, the turns it makes, and how it decelerates. Eventually, with regularity and uneventfulness, the self-driving vehicle would become just a bus that I take every day at 4pm to take me to statistics class. Many of the passengers in the pilots today have reported experiencing a feeling of boredom after a while (this is a good thing).
Although the shuttle is bounded to the university, instead of an urban free-floating environment, the service validates assumptions regarding their customers. There are people either getting in the vehicle daily or interacting with it from the outside. It is not a controlled experiment at a garage or a hardware-in-the- loop OEM test. If you’re on campus, there may be a possibility that you randomly interact with the shuttle too.
We already see in the industry efforts to improve autonomous explainability and communication to the outside world. For instance, the startup Drive.ai has adopted a digital screen on the front of their Texas City shuttle. The driverless vehicle engages with the pedestrian once it reaches an intersection and communicates that it’s safe to cross the street. In the same manner, to ease passenger anxiety about the path being driven, visible maneuvering information is already being implemented in Uber pilot programs. In the industry, this has received the name of Human Machine Interface (HMI).
Across use cases, we need to think about the reaction of the passenger and the pedestrian once they interact with the self-driving vehicle. OEMs and TNCs must go up the value chain and revisit the design phase to ensure intelligent cars are manufactured with real people in mind. Would something as an automated steering wheel, and future lack of, stir up the public? With regulation in place, would we be better off without this constant reminder of the control that has been taken away from our hands?
Technology developers must work with the underlying principle of understanding what makes [skeptical] passengers tick vs. what eases their mind. After all, the customer will have the last word.
***
I hope you enjoyed this post related to my book, Autonomousity: Autonomous Vehicles & Emerging Business Models. You can check it out via this link: https://www.amazon.com/dp/B07QDM7HTX
I’d love to continue the conversation about self-driving cars! You can either leave a comment in my Medium page or connect with me via email at BejaranoAPaula@gmail.com or LinkedIn.
