SiLC’s 4D Vision System Brings Kilometer-Range Drone Tracking into Focus

SiLC Technologies (Monrovia, CA, USA) has launched the Eyeonic Vista 4D vision system. The frequency-modulated continuous wave LiDAR based system is designed for long-range perimeter and border security/surveillance applications at ports, airports, and other such facilities and venues, says Arlon Martin, vice president of market development for SiLC. 

Designed to be robust and compact, the system can detect, identify, and track objects, including smaller objects such as drones, in a variety of environments from distances more than one kilometer away. It is also useful for detecting and tracking drones. 

From Photons to Point Clouds: System Design Explained 

Martin says that the system is a self-contained unit, usually mounted on a tilt panel, and attached via Ethernet cable to a laptop computer. The core of the system is one or more silicon photonics chips, designed by SiLC, which are fully integrated with laser, coherent photonic sensing and passive circuits, all of which function together to provide significantly better resolution than radar.   

Eight-Beam Scanning Delivers Dense, Data-Rich Point Clouds 

The unit beams 1550 nm wavelength light through eight channels that are aligned with a collimator. Each pixel provides location, range, speed and direction, and both polarization modes. A scanner is used to scan the field of view with the eight beams of light. Signal processing inside the vision system creates a point cloud which is transmitted over ethernet to a computer, Martin says. 

All of the signal processing software is developed in-house, Martin says. The computer post processes the data and performs subsequent analysis. Decisions as to what to do about the detected object are currently made by human operators can then be made by human operators or AI agents. 

Wide-Area Surveillance with Zoom-In Precision 

The field of view for the system is 30 x 60 x 30 degrees, Martin says. So, at a kilometer, for example, a field of view is slightly more than a kilometer wide and a little more than half a kilometer high. 

“That's a very large field of view if you're looking over, for example, a stadium or an airport,” he says. “So, when you see something, which might be a drone or it might be a bird, that you want to have a closer look at, then you scale it down to have a very short field of view, down to a range of around 20 x 40 meters, to get more of your visibility, just like a pair of binoculars or another camera. Now you can really zero in on what it is that you want to look at and get a very high-resolution view. We can see objects that are less than half a foot in size, so that gives us a very good, very high resolution at those ranges.” 

Centimeter-Level Accuracy at Kilometer Distances

One unique aspect of the system, Martin notes, is that while it is scanning and sending the pixels back to the vision system, it can measure the distance, with high accuracy, of every pixel. 

“That means we know exactly how far away that object is and the speed of it,” he says. “Our accuracy of measurement at a kilometer is around 10 centimeters. Radar can't do this; a camera can't do this; a camera doesn't know how far away objects are. Even with a range finder, they just give you a general idea.”

Knowing the exact distance of an object is critical to making decisions for taking or not taking action, Martin notes. For example, if an object turns out to be a hostile drone, a decision may have to be made to ground or destroy it. 

“We cannot only tell you how far away it is, but which direction it's going and how fast,” he says. “With every pixel, we provide that velocity information. And that's why this is a very powerful, unique vision system that just didn't exist before. If you see something, you can focus on it. Maybe you see three or four objects if it's a swarm. And in that case, since you know the velocity of the different objects, you might then prioritize the action to take with each one.”

One challenge radar has is that it will lose track of a drone if it hovers, moves slowly, flies against a busy background such as a heavy tree line, or drops below the horizon. By contrast, this system can track a drone, even if it is motionless, Martin says. 

“We actually see, in many cases, the propellers of the drone rotating,” Martin says. “Because we do velocity detection, we know how fast and where it is going, and if it stopped, we know where it is. We operate with invisible light, and we operate day or night, so glare or low light does not matter; it makes no difference for us.

Another advantage, Martin says, is the system’s capability to distinguish between objects of concern, like drones, and friendly objects, such as birds. That’s because the system looks at the velocity signature of an object in addition to the polarization; it can tell the difference between propellers rotating and wings flapping.

“These are things that radar and cameras have a hard time doing at long distances,” he says. “We can do that because we detect both polarization modes of light.”

Long-Range Sensing Comes with Signal Tradeoffs

One challenge for the system is the fact that long distance scanning and detection means working with weaker return signals. That means target reflectivity and harsh environmental conditions such as fog, rain, or dust can have an impact on range and image quality. 

“Maybe we're seeing things at a kilometer-and-a-half on a on a clear day and then a foggy day...it's at a kilometer,” Martin says. “We've done dust tests, we've done fog tests, and all of the images do degrade, but we have a whole lot better visibility than we do with the human eye or with other sensors, camera sensors in particular,” Martin says. “That's partially because we can do things like range gating, because we know the distance of every pixel.”  This means that with the aid of AI, much of the noise from fog, dust, or similar environmental interference can be reduced, allowing the system to focus on the target object. While the resulting image may not appear as clean, the system can still detect and track objects—such as a drone—even through cloud cover.

Related: Bitsensing Launches 4D Imagin Radar for Autonomous Vehicles