Echodyne Targets Drones, Self-Driving Cars with Metamaterials Radar
Echodyne, a secretive Seattle-area startup company backed by investors including Bill Gates and Paul Allen, is developing a novel, high-performance radar suitable for drones, robots, and self-driving cars. The technology could potentially allow such vehicles to operate independently in a range of conditions.
The company, housed in a drab, unmarked building just off State Route 520 in Bellevue, WA, thinks it can dramatically improve upon current radar systems in terms of cost, size, weight, and performance, by using metamaterials, which Echodyne co-founder and chief technology officer Tom Driscoll describes as “sub-wave length geometric configurations of metal and circuit board.” (We’ll dive into that in a minute—for now, think tiny structures that can change the way a surface interacts with radio waves.)
Echodyne will be marketing its technology to government and military customers, the traditional behemoths of the radar industry. But its co-founders are more excited about building a commercial business serving new markets and applications that haven’t used radar before because it was too expensive, too heavy, or didn’t offer a meaningful improvement over existing optical sensing technologies.
“We have this concept of radar vision, where you’re actually using radar as a vision system for autonomous and unmanned vehicles as opposed to an exotic military-grade only sensor,” says co-founder and CEO Eben Frankenberg.
Echodyne is the most recent metamaterials company to spin out of Intellectual Ventures, the patent licensing and invention business co-founded by former Microsoft chief technology officer Nathan Myhrvold. Intellectual Ventures amassed a significant metamaterials intellectual property portfolio and further developed the technology through its Metamaterials Commercialization Center, which Driscoll directed.
The other companies are Kymeta, which has raised $82 million to apply metamaterials technology to satellite communications (and Monday announced a partnership with Airbus Defense and Space focused on the maritime industry), and Evolv Technology, based in Boston and funded to the tune of $12 million to tackle advanced imaging.
“There’s a perception that because [neither] Kymeta, nor Evolv, nor us have launched any true product yet, that all of metamaterials is being squirreled away in some military domain and no one will ever see it,” Driscoll says. “It’s not our intent. We will be partnering with military. We will also be making commercial products.”
To demonstrate this, Echodyne installed a prototype metamaterials radar antenna—which looks like a stack of printed circuit boards, about the size of a shoebox top—on an off-the-shelf quad copter. It’s capable of lifting a little more than six pounds, though the prototype unit weighs only about 2.6 pounds. The company modified the flight controls so that the drone could autonomously follow a target, using the radar antenna to guide it. Frankenberg and Driscoll say the drone—which they showed off on a conference table, but did not allow me to photograph—followed them around a field.
“We gave a tiny quad copter the ability to see and image the world around it with radar,” Driscoll says, calling it a first for a radar system capable of scanning for new objects while tracking existing ones—capabilities previously confined to high-end electronically scanning radars.
“You couldn’t even lift anybody else’s electronically scanning radar with this [quad copter] at almost any price,” Frankenberg says.
To explain why Echodyne believes it can upend the status quo, Frankenberg and Driscoll laid out the limitations of current radar technologies—and the advantages of metamaterials.
Radar, an acronym dating from the 1940s for radio detection and ranging, was demonstrated in rudimentary form at the beginning of the last century. It works by sending out a pulse of radio waves, which bounce off objects in the way. By measuring how long it takes the radar signal to return to the radar antenna, you can determine the presence and position of distant objects.
In the familiar mechanically scanned radars, a single beam is transmitted from a rotating antenna to detect objects in a wide area. That’s the main technology available for commercial applications today, Frankenberg says. Entry-level models for boats cost in the low thousands of dollars.
More modern radars direct the beam using phase shifters at each antenna element to create a radio wave front that travels in a desired direction. Frankenberg says the state of the art today is the active electronically scanned array, in which many individual antenna elements work in concert, allowing the radar beam to be pointed almost instantly in any direction—and to switch quickly between scanning for new objects and tracking existing targets. These high-performance systems start at $100,000 and go up into the millions of dollars each, he says.
Mechanical radars can only do so much, while electronically scanned arrays remain expensive because of the phase-shifters and amplifiers placed at each antenna element. “The only way to drive the price down is to somehow make the electronics behind every one of those antenna elements cheaper, through volume, but there was no physics change whatsoever,” Frankenberg says.
Here’s Echodyne’s big advance: “With our metamaterials approach,” he says, “we fundamentally change the physics of the way the antenna works. That allows us to make a huge improvement in cost, size, weight, and power.”
That improvement, Echodyne believes, will unlock huge new markets. “I like to compare this in some ways to the Global Positioning System, which started out as a high-end, primarily military sensor and is now on countless devices,” Frankenberg says.
Metamaterials off the shelf
To the layperson, metamaterials—like any technology insufficiently understood—start to approach magic. Just to be clear, I’m a layperson, but Frankenberg and Driscoll did their best to explain the technology, while keeping trade secrets close to the vest.
“When we use the term ‘metamaterials,’ people often think this a materials science project, and you’re building some kind of exotic material to do this,” Frankenberg says. (Some very high frequency applications do require exotic materials, but not the ones Echodyne is working on.) “It’s printed circuit board technology, with pretty standard off-the-shelf electronics parts,” he says.
The geometric configurations of metal and circuit board—gold-colored swirls to the naked eye—that form the metamaterials surface act as the individual antenna elements or cells in Echodyne’s radar.
“Different people have found that different shapes are more or less advantageous in different configurations, and we’ve found the same thing,” Driscoll says. “So part of our secret sauce is in what shapes work best for what radar applications, and how you build those shapes into circuit boards.”
The metamaterials antenna elements don’t need individual phase shifters or amplifiers like in active electronically scanned arrays. They’re controlled instead by “fairly simple row-column addressing software” that turns them on and off, creating patterns to direct the radar beam, Frankenberg says. This saves costs both in individual electronics components and cooling systems required to keep them from overheating. It also allows more antenna elements to be placed closer together for higher-resolution radar.
A useful analogy can be found in video monitors. “If you get up close to it, you can see pixels,” Driscoll says. “When you step away from it a little ways, they blend together. They homogenize to form an image. That’s exactly the same thing we’re doing here.”
Bottom line, Echodyne says the metamaterials antennas will offer capabilities like active electronically scanning arrays at prices closer to mechanical radars, and in much smaller packages.
The company has 13 employees and about half a dozen job openings, with plans for more as it ramps up research and development, engineering, and manufacturing of its radars. The company aims to have two products at different frequencies ready this year to share with integration partners, which would combine Echodyne’s antennas with their existing back-end systems for radar signal processing.
Radar for cars
Echodyne is positioning its metamaterials radar among other machine vision technologies, such as optical cameras, near-infrared and structured light sensors, and Lidar (a radar-like technology that uses laser beams in place of radio waves), that will enable a new generation of autonomous vehicles to operate safely in a normal outdoor environment.
In addition to aerial drones and new military and government applications, Echodyne sees great long-term potential for metamaterials radar in the automotive industry. While low-end radar is used now for things like adaptive cruise control or parallel parking assistance, “we think the value proposition for a true imaging radar is significantly higher than that,” Driscoll says.
The company believes radar is a superior sensor because it can accurately detect objects and the distance to them, while optical systems may require multiple sensors or sophisticated software to do that. And radar, unlike other machine vision technologies, works in rain, snow, darkness, dust, and other low-visibility conditions. “Optical systems all start to fall apart when you introduce any kind of environmental variables,” Frankenberg says.
Indeed, one of the initial uses of radar was to find ships in the fog. In a few years, perhaps, it could be used to help driverless cars better navigate foggy streets.