Montana company stands out in the lidar crowd

A technology first developed to assist the U.S. military in spotting enemies sneaking past battlefield perimeters could soon help self-driving vehicles create detailed portraits of their surroundings.

From its roots in supplying the Army, Air Force and Marines with lidar sensors used for monitoring territory, Blackmore, an imaging company based in Bozeman, Mont., has spent the past two years refining its technology for use in autonomous vehicles.

Blackmore is hardly alone in that pursuit — industry experts are tracking at least 60 lidar companies all vying for business as automakers and major suppliers formulate production plans for autonomous-driving systems. Sensors that create a high-definition depiction of the road ahead are a critical part of those systems.

Amid that crowded field, Blackmore stands out for an unconventional approach its founders say results in a key advantage.

In addition to providing detailed information on the distance to objects, the company's technology can also measure the velocity of those objects, information that conventional lidar companies usually cannot provide. For autonomous vehicles attempting to make sense of their road environment and plot a course ahead, that insight on the speed of other actors in the space can be critical.

Competitors downplay the potential advantage of Blackmore's approach, though, saying it involves making tradeoffs in range, cost and complexity.

Though their prices are collectively falling, lidar units are expensive, ranging from $1,000 a unit to more than $100,000 for sensors the size of a hockey puck. So wringing more information from them helps justify their cost.

"The prices haven't dropped yet, resolution is poor, so we ask ourselves, 'What other data can we get off these things?' " says Gill Pratt, CEO at the Toyota Research Institute. "Getting the velocity matters a whole lot."

Toyota AI Ventures is an early investor in Blackmore, as is BMW i Ventures. Collectively, Blackmore has raised $21.5 million through two funding rounds, per Crunchbase data. That's a sliver of the $1.1 billion that investors have poured into lidar companies in recent years, according to CB Insights, a consulting and research firm.

Lidar sensors emit lasers to detect and measure the distance to objects in their field of view. Most companies modulate the amplitude of pulsed wave signals — similar to AM radio — and measure the time it takes for a light wave to reach an object and bounce back.

Blackmore, on the other hand, has zagged where others zig. It uses frequency modulation — akin to FM radio — to make its measurements. Rather than a pulsed signal, its lidar sends a continuous wave in which subtle changes in the frequency can be made. Similar to radar, this allows the lidar unit to measure the Doppler effect, the change in wavelength as distance between a unit and object changes.

Much like many radio broadcasters switched from the AM to FM band over decades, Randy Reibel, CEO of Blackmore, believes lidar sensors will make a similar transition.

"Physics doesn't respond to money," he said. "FM lidar may be a little behind AM lidar, which has been around for 25 years. But when you look at scalability and differentiation, the long game is FM. Systems trend toward FM. Our company firmly believes that. I've spent 15 years looking at these trends, and it's taken us a long time to lay the roots and foundations for what we're doing."

Blackmore will take another step in that evolution this year, when it ships its second-generation Autonomous Fleet Doppler Lidar units. They're geared toward the robotaxi market, and are scheduled to reach customers in the second quarter.

Depictions of the Doppler effect were clear during demonstrations of the new units at CES last month during ride-alongs in Las Vegas, with Blackmore's sensors collecting enough information to show tires rotating on vehicles 200 yards away, as well as granular information that depicted the limbs of pedestrians.

That's a direct outgrowth from Blackmore's military history, in which the company pioneered the frequency-modulated lidar approach enough so that its analytics software could spot hand movements of people detected in forward operating areas, determine what objects they were likely carrying, and whether they were friendly or hostile.

In the autonomous-vehicle realm, lidar sensors do not, of course, solve the perception challenge all by themselves.

Data from lidar is fused with information from radar and cameras, and in some cases infrared and ultrasonic sensors are also utilized in the sensor configuration. Velocity information can already be determined with radar, in which case some might view the ability to achieve the same result with lidar as redundant. Reibel disagrees.

"In a sense, this covers a base that radar has traditionally and uniquely covered," he said. "But in a way, that's like asking, 'Would you prefer a seatbelt or an airbag?' I would like both, and we think it pairs better in a sensor-fusion context with radar and cameras also trying to detect motion."

Velocity information isn't the only advantage conferred by a frequency modulation, its proponents say. Because the light wave is continuous, sensors using frequency modulation don't require as much power. If AM-based lidar sensors need the equivalent of a bright flashlight to see a certain distance, FM lidar might need only a dim flashlight to accurately see the same distance.

Blackmore is no longer the only lidar company which sees the potential benefits of frequency modulation. Last year, Aeva, a company founded by two former Apple engineers, showcased its first lidar sensor based on frequency modulation. It would be easy to assume Blackmore executives would feel unnerved by the new competition, but that's not the case.

"I feel great about that," said Jim Curry, vice president of analytics at Blackmore. "It has stirred up a productive conversation about the merits of AM versus FM, and the fact that there's more than one player pushing this technology is good for the lidar industry. It was starting to get a little lonely. Now people are asking questions, and they're forced to justify their approach."