Published on December 19th, 2018 | by Steve Hanley0
Autonomy News — December 2018 Edition: Nuro, Formant, & Anthony Levandowski
December 19th, 2018 by Steve Hanley
The convergence between humans and machines continues gaining speed as the end of 2018 approaches. We often associate autonomy with vehicles but it can apply to virtually any aspect of human endeavor from mixing concrete to making movies. Machines never get tired or bored. They don’t come into work hungover on Monday or have to stop every 5 minutes to share a selfie of themselves with 10 million of their closest friends.
Anything a human can do, a machine can do better. That’s the theory, anyway. In practice, machines still need human input every once in a while, especially when they find themselves in unusual situations they have not been programmed to handle. But the number of such instances is getting less every day. Here are three examples.
Nuro Driverless Grocery Deliveries
Self-driving startup Nuro has been using specially modified Toyota Prius automobiles to deliver groceries to Kroger customers for some time now, but always with a human safety driver along. Now it has begun using its R1 fully autonomous delivery vehicle. The R1 looks a little like the cute Google car from a few years ago with a carrying handle grafted on.
In development since 2016, the R1 has two compartments that can hold up to 6 standard size grocery bags each. “Nuro envisions a world without errands, where everything is on-demand and can be delivered affordably,” president Dave Ferguson says. “Operating a delivery service using our custom unmanned vehicles is an important first step toward that goal.”
According to TechCrunch, Nuro intends to use its self-driving technology for last mile deliveries of local goods and services. It could be groceries or dry cleaning, but it applies equally well to any items that can fit inside one of the compartments in the R1. As long as the pickup and delivery points are within the defined area the vehicle is programmed to operate in, the R1 can handle any delivery chore large or small.
Formant co-founder and CEO Jeff Linnell says, “What’s really going to move the needle in the innovation economy is using humans as an empowering element in automation.” According to TechCrunch, the idea behind Formant is to equip businesses with cloud infrastructure for collecting, making sense of, and acting on data from fleets of robots. It allows a single human to oversee 10, 20, or 100 machines, stepping in to clear confusion when they aren’t sure what to do.
“The tooling is 10 years behind the web,” Linnell says. “If you build a data company today, you’ll use AWS or Google Cloud, but that simply doesn’t exist for robotics. We’re building that layer.” Linnett began his career in robotics by designing the special dollies that made the things and people in the movie Gravity appear to be weightless. He formed a company to market the technology, which was later purchased by Google. He then spent 4 years at Google as a director of robotics before leaving to start Formant with chief technology officer Anthony Jules.
First, says TechCrunch, Formant connects to sensors to fill up a cloud with LiDAR, depth imagery, video, photos, log files, metrics, motor torques and scalar values. The software parses that data and when something goes wrong or the system isn’t sure how to move forward, Formant alerts the human “foreman” that they need to intervene. It can monitor the fleet, sniff out the source of errors, and suggest options for what to do next. Formant’s algorithms learn from every such human intervention so next time the machine encounters a similar situation, it won’t need assistance from a real person.
Linnell says he worries about people or companies who try to do similar things without the proper background. “Putting robots online in an insecure way is a pretty bad problem.” His company is rushing to identify and resolve any security issues before it begins distributing its software to customers next year.
With time, humans will become less and less necessary, and that will create enormous societal challenges for employment and human welfare. “It’s in some ways a continuation of the industrial revolution,” Jules suggests. “We take some of this for granted but it’s been happening for 100 years. Photographer — that’s a profession that doesn’t exist without the machine that they use. We think that transformation will continue to happen across the workforce.”
That sunny outlook is fine but seems a bit disingenuous. Robots are going to make a lot of the jobs humans do today unnecessary, a situation envisioned by the Disney movie Wall-E a decade ago. Linnell and Jules may not have accounted for all the societal impacts the brave new world presaged by their software may entail.
Look, Ma. No Hands!
Elon Musk says one day soon, a Tesla will drive itself from LA to NYC without the aid of human driver. Serial tech bad boy Anthony Levandowski says he has already done it using a specially modified Toyota Prius. He says his journey took place in October and he claims to have time lapse video that proves his claim.
Levandowski’s ethically challenged behavior was documented in detail by Lawrence Burns, who wrote the definitive book on Google’s self-driving program entitled Autonomy: The quest to build the driverless car. In it he describes how Levandowski created his own company to compete with what Google was doing while still part of the Google team. He eventually left unannounced, taking reams of data and software with him on the way out the door.
Google sued after he started Otto, a company that made self-driving software for tractor trailers, which was soon snapped up by Uber. Levandowski strenuously denies any wrongdoing. Nevertheless, he asserted his Fifth Amendment right against self incrimination more than 500 times in depositions during the discovery phase of the Google lawsuit, according to The Guardian.
What is noteworthy about his latest self-driving Prius is an absence of Lidar sensors. Levandowski says the reason fully autonomous cars don’t exist yet is because the software technology doesn’t exist, not because Lidar technology isn’t good enough,
The system mounted in the Prius uses six video cameras pointing to the front, sides, and rear of the vehicle, each with a lower resolution than those found in modern smartphones. Images from the cameras are fed to the trunk where a computer running two neural networks is installed.
One neural net recognizes lane markings, signs, obstacles and other road users, and extracts information about their position and speed. The other takes that information and controls the driving using digital signals and mechanical actuators for the throttle, brake, and steering.
A seventh camera faces inward to make sure drivers are keeping their eyes on the road at all times. If a driver looks away, nods off, or pulls out a cellphone, the system sounds alerts that become louder and more persistent until ultimately the system brings the vehicle to a halt. The Guardian says it witnessed the system in operation.
Levandowski has formed a new company called Pronto.ai to market the self-driving system, called Copilot, to trucking firms. The basic installation costs $5,000. It is intended for highway use only and is not capable of driving in urban environments where it has to contend with cross traffic, pedestrians, bicyclists, and a thousand other distractions. “There are more self-driving scenarios that we need to handle than there are atoms in the universe,” Levandowski says.
More than 4000 highway deaths involving heavy trucks happen in America every year. Ognen Stojanovski, a lawyer and research scholar at Stanford University who co-founded Pronto.ai, says, “Trucking is a tight margin business. Driver retention is a huge cost, and if we can add even a little bit of safety, lower claims from less severe crashes will make a huge difference.”
“If true, a truck that used only cameras to steer, brake, and accelerate for 100% of any cross-country trip is impressive,” Bryant Walker Smith, a law professor at the University of South Carolina told The Guardian. “Making a system work with cameras alone could be a major contribution, especially if this could be applied to higher levels of driving automation.”
Notice the phrase “If true.” Missy Cummings, director of the Humans and Autonomy Laboratory at Duke University, is skeptical. “Anthony’s job is to make claims that may be at the edge of what his technology is capable of,” she says. “I have not seen evidence of amazing breakthroughs that would be a game changer in driverless car technology, particularly if it’s only relying on cameras.”
The Guardian also spoke to the heads of two other autonomous driving companies — neither of which agreed to be identified. Levandowski and litigation go together like baseball and hot dogs. One said, “The real test is how repeatable it is.” The other added that Levandowski is still “radioactive” in the industry, which may make it difficult for him to attract investors.
“I’ve learned a lot in the last couple of years about how to do engineering, both on the technical side but also how to operate and be more responsive to people’s criticisms,” says Levandowski. “We’re not promising the moon. We want to promise things that are very concrete and that we can deliver.”
Elon Musk and his Tesla Semi team may have a thing or two to say about self-driving trucks in the near future, and Levandowski’s old pals at Waymo are also field testing their own autonomous driving software for heavy trucks. Does Pronto.ai have what it takes to compete in what is shaping up to be a crowded market? “We’ll see,” said the Zen master.
Have a tip for CleanTechnica, want to advertise, or want to suggest a guest for our CleanTech Talk podcast? Contact us here.