In The News
Technori: Navigation is stuck in the 20th century. Here’s what’s coming next.
Planning a road trip used to mean digging out the crumpled Rand McNally atlas from your trunk, or printing out turn-by-turn directions from MapQuest.
If you were right on the cutting edge in the middle aughts, you might have had a sat nav that occasionally guided you into large bodies of water.
Google Maps debuted in 2005 and changed the game. Once smartphones became commonplace, it seemed like our understanding of directions had peaked: we would never be lost again.
But one man didn’t buy into that.
Even before the rest of us were blown away by the Maps app, David Mindell was working on delivering even more precise location information. In addition to his various appointments in aeronautics and robotics at MIT, David has been involved in dozens of underwater research explorations, including the search for the Titanic. It was the navigational challenges that come with being underwater — where GPS can’t reach you — that inspired him to find a better way to help everyone figure out exactly where we are at all times, by mapping every single section of space, down to the cubic millimeter.
This isn’t just for humans: David and his company Humatics are focused on developing technology that will help robots better understand where they are, which can help make their movements more precise, and their interactions with other robots and humans more efficient and even safer.
In the interview, David explains how navigating the deep oceans taught him the importance of knowing where you are, and what it’s like to be 20 years ahead of everyone else.
Scott: Could you explain what Humatics is, and why you decided to try to solve this problem?
David: I’ve spent my career doing both engineering and the humanities. I grew up in the seventies and eighties, was heavily influenced by the first wave of personal computers, so ever since Steve Jobs with the Apple II, I always had my eye on becoming an entrepreneur. My mother is an entrepreneur, my father is an engineer.
I developed precision navigation technology in the eighties and nineties when I was working in deep ocean robotics. I worked with Bob Ballard and the group that found the Titanic: we spent 15 years traveling around the world looking at shipwrecks and hydrothermal vents and other things in the deep ocean. I got hooked on precision navigation and navigating robots to the centimeter — and even more accurately. I licensed that technology off twice, and that got me convinced that there was a market for precision navigation in the terrestrial world. Everybody became enamored with GPS and what it can do, but it’s also very limited in that it doesn’t work indoors, it doesn’t work underground, and it doesn’t work in cities. And it’s not very accurate: it’s 10 to 30 meters of accuracy, depending on the conditions.
I went to MIT, did my Ph.D., and had a very nice life as a professor for 20 years. But this idea of starting a company to do precision navigation never left me. I tinkered around in my workshop for about five years, teaching myself microwave and radar electronics — which is a different kind of electronics than I specialized in in the robotics world — and I learned enough to know that it was possible.
I started talking to people and nosing around. This was 2014, 2015. There were a lot more robots coming online. I was working in autonomous systems and building autonomous helicopters and other types of research, and you could see that the robotics world was about to explode. Driverless cars were on people’s minds. And navigation is key to any of these technologies.
I was a sort of insider/outsider in the technology world. I wrote five books about the history of technology, and about autonomous systems and robotics, so I was used to looking at the world from the outside as an analyst; and then as an engineer I was interested in building systems that worked in the real world, with all of the constraints that puts on what you can accomplish, as opposed to working in a research lab.
Thrown in at the deep end
Scott: What do you think it is about your background or the way your brain works that enabled you to identify the application versus just the novelty?
David: Working undersea, we had to invent our own navigation. Even then GPS was pretty new, and it works underwater but only to a depth of about a millimeter — and we were working in 10,000 meters of water! You have to invent your own navigation system. I learned a lot about navigation, and that more accuracy than anyone ever thought they needed would always be useful, and people would eventually become addicted to highly accurate navigation.
In the underwater world, we learned things that were a decade or two ahead of people in other environments. I’ve always been good at translating knowledge across those kinds of boundaries: it teaches you to see things in context, and to make connections between things that people don’t otherwise think of as connected.
One way to describe Humatics is, we’re a robotics company that doesn’t make robots. We make the connective tissue that connects robots to the world by telling them where they are and how to navigate. We combine radar and radio frequency and ultra-wideband engineering with robotic navigation: that’s a unique combination, there’s no other company out there that does that. And that enables us to build systems that make robots incredibly productive and useful and robust in the world.
Navigating the future
Scott: In Humatics, this entire career that you’ve had is all coming together in one. It’s been in your mind for 30 years, but here we are just entering the real first base of robotics. Does it blow your mind that this all seems to continue unfolding like it was meant to be? Or do you feel like you’ve been sitting here waiting?
David: I really did plan to start this company for 20 years — and then one day it was the time. During that time, I was building my knowledge, I was building my network, I was building my understanding of what technologies are out there. I was always interested in robotics because it was the way that technology and software interacted with the physical world. Very difficult and very challenging, but very interesting.
The physical world is still not interactive with the digital world. GPS is a very important first step in that, and that enables all the things we take for granted, like Google Maps and Uber; but it’s only one way to think about navigation. In fact, I think it’s a very 20th century idea — that you have a single coordinate frame that covers the entire globe. It’s a very valuable thing, obviously, when you’re traveling outdoors and long distance. But it doesn’t work indoors, underground, or in the built environment that we increasingly live and work in.
When you work in those environments, the single coordinate frame that covers the world is not the right answer. The right answer is a local coordinate frame, like the coordinate frame of the factory, or the coordinate frame of your loading dock where you’re just trying to get a job done and you need to do it with very high precision, and you don’t care where you are on the surface of the Earth. You care where you are relative to the ship, or the crane, or another robot. That introduces the idea of navigation as a relationship: rather than a single coordinate frame that covers the entire Earth, you have many thousands or millions of coordinate frames that have this fractal growth across the landscape. They basically are the connective tissue that tie a robot into the building that it’s working inside, that tie a train into where it is very precisely within the station it’s trying to track into.
The whole idea of navigational relationships is really what the company is founded on: that’s the key that’s going to bring the digital world into broader contact with the physical world. Look in front of your face right now, imagine every little cubic millimeter: we’ll soon have a digital tag attached to it.
People say, what’s your vision for five years from now, and I say, five years from now, we’ll say, can you believe that in 2019 we didn’t know where anything was? It’s one of the oldest questions of humankind: where am I? There’s a lot of complex technology, but at the end of the day, it’s a question that everybody can relate to and everybody needs the answer to. If I know where I am in relation to the things around me, I can begin to live in a rich digital-physical environment.