At 1:30 in the morning on Saturday, June 17, the ACX Crystal, a 730 foot long cargo ship weighing 29,000 tons and carrying nearly 1,000 containers, plowed into the side of the USS Fitzgerald, a US Navy destroyer based at the Yokosuka naval base in Japan. Is there a connection between what happened and autonomous driving technology? Maybe more than most people realize.
Death At Sea
The Fitzgerald, weighing only 9,000 tons, got the worst of the encounter. The collision killed 7 members of the crew who were sleeping in their bunks below decks. The bulbous bow of the ACX Crystal penetrated the hull of the Fitzgerald below the waterline. The crew members drowned when the decision was made to close watertight doors to keep the destroyer from sinking. That decision trapped them inside the flooded compartments.
Ship tracking websites show the ACX Crystal altered its course only slightly after the collision, then proceeded on toward Tokyo as if nothing had happened. It was not until 30 minutes later that the container ship made a 180º turn and came back to the vicinity of the collision.
Like A Computer Was Driving
Steffan Watkins is an information technology security consultant. He writes for Janes Intelligence and specializes in tracking the path of ships. He tells the New York Times the course as revealed by the Crystal’s Automatic Identification System “looks like an automated course.” To clarify: “It looks very much like the computer was driving.”
And what of the Fitzgerald? Destroyers are by definition fast, highly maneuverable vessels that should be able to sprint away from danger. Standard operating procedures on most naval vessels require lookouts with binoculars be stationed on both sides of the ship and at the stern, constantly sweeping the horizon for signs of danger.
In addition, radar operators on the ship’s bridge and in the combat information center should have seen the ACX Crystal approaching miles away. That’s a total of 5 people who failed to see what should have been seen on a clear night on a calm ocean. Does that raise the possibility that a computer was driving the Fitzgerald, too?
Crew members of the Fitzgerald have been ordered not to speak to the press while the investigation into the collision and subsequent fatalities is underway, but one person aboard managed to get this message to the New York Times via social media: “All I can say is, somebody wasn’t paying attention.”
Computers & Human Drivers
The link between the collision at sea and autonomous driving vehicles is the still imperfect relationship between computers and humans. Sebastian Anthony, writing in Extreme Tech, says, “For some reason, we often think of computers as infallible — subjective, logical, rational, and nearly always right. There is something about a computer’s lack of emotion and intelligence that makes them strangely trustworthy.”
Engineers working on autonomous driving systems are challenged by what is known as the “handoff phase.” That is the moment in time when a computer says, “Whoa. This is way too much information for me to handle. I’m going to give control back to a human driver to figure out what to do next.”
The handoff period may take a second or less. A person reading the Wall Street Journal, playing solitaire on a digital device, or taking a selfie while driving may need 10 seconds or more to refocus on the road ahead, figure out what the best course of action is, and then act on that information. During that interim, the chances of a bad result ending in injury or death are greatly increased.
Autonomous Driving Risks
The wreck of the USS Fitzgerald is a tragedy for the families of those who died. It is also a cautionary tale for those who are charging full speed ahead into an era in which autonomous driving cars are as common as self-service elevators. Surveys show not everyone is comfortable with that idea, with older drivers more likely to be wary of too much technology.
The implications for autonomous driving cars are clear. Even if computer systems can keep a car in its lane on the highway, complex situations involving traffic lights, stop signs, pedestrians, bicyclists, and stray dogs will continue to be a daunting challenge. No doubt, mankind will get there and there is every reason to believe Elon Musk and his band of merry computer nerds will get there firstest with the mostest.
Musk has made autonomous driving into a crusade. He is not one to be easily deterred. In fact, he is now on his 4th head of the Autopilot program in the last 7 months. He intends to continue his quest until he can find someone who will give him what he wants. Autopilot is very sexy and garners lots of media attention, but it may come at the expense of other basic human necessities with a higher social value.
In the meantime, what happened in the Sea of Japan a week ago should put us all on notice that the machine/human interface is still an area beset with difficult problems. The new Tesla Model 3 will have all the hardware it needs to drive by itself from coast to coast, but the software to make that possible?
Congress is considering legislation to establish nationwide standards for autonomous driving technology. Perhaps what happened to the USS Fitzgerald should be a cautionary tale for those who say we should move ahead as quickly as possible and fix any issues that crop up later.
Source: New York Times
Don't want to miss a cleantech story? Sign up for daily news updates from CleanTechnica on email. Or follow us on Google News!
Have a tip for CleanTechnica, want to advertise, or want to suggest a guest for our CleanTech Talk podcast? Contact us here.