A man who knows where he is

Autonomy for a machine or a man is never possible because we're always interacting with an environment.

By Kevin Parker May 5, 2018

David Mindell is CEO and co-founder of Humatics, a company changing the ways people and machines locate, navigate, and collaborate. His company is among those enabling greater use of robots in industrial environments.

Mindell is also an MIT professor of aeronautics and astronautics, and of the history of engineering and manufacturing. He’s author of five books, including Our Robots, Ourselves: Robotics and the Myths of Autonomy. Besides being an electrical engineer, he has a Yale literature degree, which no doubt contributes to the excellent quality of his prose. It must also be mentioned that he’s been involved in more than 25 oceanographic expeditions.

"The myths of autonomy" alludes to the fact that, whether for a human or machine, autonomy is never fully achieved. We are always interacting with an environment. Therefore, in industrial environments and in all of life really, "Micro-location is the substrate for greater collaboration," said Mindell. For, as all of us have discovered at one time or another, we’re not as sure of where everything is as we often assume. 

Micro-locations in millimeters

For a robot coming up to a workstation, millimeter precision as to location may be needed, with point-to-point measuring capability using radio-frequency technology. That ability is part of Humatics’ intellectual property. One scale up is centimeter precision, and the one above that is equivalent to that of a GPS, where 6-meter precision is possible at best.

"Software brings the levels together," Mindell said. "So that the machine jumps from one level to another, going to the source of the best scale at the time, which always involves multiple sensors and an algorithm that makes the judgement."

Mindell suggested the location task is like driving a car under different road and weather conditions. "The headlamps need robustness, for dealing with things like snow or fog. The streetlight gives you a location fix."

More generally, as Mindell explains in a video available on YouTube, the highest level of automation doesn’t strive for autonomy. It seeks something more complex and difficult, but that delivers greater financial and social benefits. It might be thought of, Mindell said, as a wrapper of human intentionality.

The question, then, isn’t really whether some robot will take it upon herself to run amok. Rather, what were the intentions and desires of those who infused what intentions and desires into the robot? 

Myth list

In any case, a lot of engineering systems are launched with the idea of full autonomy in mind, Mindell notes, but somewhere in the movement from lab to field a mix is come upon. "Automation transforms human interaction but doesn’t eliminate it. It’s possible to trace any network, or see any computer programming task, as the placing of human intention, human design, and human viewpoints for storage and later execution."

The most sophisticated autonomous machines are developed for extreme environments, and in his book, Mindell examines their use in deep ocean, high atmosphere, and outer space, to see what they can tell us about how we’ll interact with cars and in daily life in the future.

Bring rigor to thought

Another author graces this edition of IIoT for Engineers. Jeff Kavanaugh is a senior partner at Infosts. He is also an adjunct professor at the University of Texas at Dallas. His book Consulting Essentials was published in April 2018.

It’s important that human thought, even in an era of computerization, have sufficient rigor. To improve critical thinking, anyone can follow the same four-step process Kavanaugh says the world’s top consultants use to manage their project.

Kevin Parker, senior contributing editor, CFE Media, kparker@cfemedia.com.

This article appears in the IIoT for Engineers supplement for Control Engineering and Plant Engineering

See other articles from the supplement below.