Insight Photonic Solutions
A driver’s ability to see and avoid other vehicles is essential to safe driving. The same is true for autonomous vehicles, which rely on sensors to allow them to navigate safely through traffic. A technology that allows ophthalmologists to study the retina of the eye is being considered for the latest generation of autonomous vehicles.
Autonomous vehicles employ various sensor types to perceive and understand the world around them including cameras, radars, ultrasound, and LiDAR (Light Detection and Ranging). Each type has its own strengths and drawbacks for this application. LiDAR sensors offer the advantage of providing detailed 3D images but in the past, their range has been limited to distances of roughly 80-100 meters or less. The LiDAR sensors that have been deployed so far in test vehicles are based on a time-based technique known as Time of Flight (ToF) or direct detection. Frequency Modulated Continuous Wave (FMCW) is an alternative LiDAR technology that offers many advantages over ToF, including higher sensitivity, immunity to stray light, and the ability to detect velocity directly.
One major difference between ToF LiDAR and FMCW LiDAR is their laser source requirements. ToF employs a simple, relatively inexpensive pulsed laser diode; however, FMCW demands far more precise control over a tunable continuous wave laser source in order to sweep it through its wavelength chirps very linearly, very fast, and very repeatably. For automotive LiDAR design, one of the biggest challenges is designing laser sources that offer sufficient coherence length to produce a beam that retains coherence long enough to reach the farthest target required and return to the sensor without breaking up.
Insight Photonic Solutions has been developing swept laser sources — from work done at UC Santa Barbara to develop tunable lasers for telecommunications applications — for a dozen years for use in Optical Coherence Tomography (OCT), which uses coherent detection to look at features of the human body for biomedical imaging and diagnostics applications. These include applications in ophthalmology, dermatology, cardiology, and more than a dozen other specialties. The Insight swept-source laser’s high measurement speed reduces movement blurring and its wide tuning range supports high-definition 3D imaging. This technology found its first commercial medical imaging application in ophthalmology and OCT is now a standard of care in that industry.
Researchers at Insight Photonic Solutions understood that OCT and FMCW LiDAR were essentially the same technique, just known by different names. Experimentation confirmed that their OCT imaging system could be used as an FMCW LiDAR system. The researchers turned a system that was being used to image a mouse brain on its side and captured an image of the lab. The team, led by Dr. Chris Wood, then modified the swept-source laser and detection system to optimize them for the longer ranges required for automotive LiDAR.
Although OCT and FMCW LiDAR share the same core coherent detection technique, further development work was necessary to adapt Insight’s technology to autonomous vehicle applications. In OCT applications, the goal is micron-level resolution to produce precise images of biological structures at close range. In automotive Li-DAR, however, the goal is imaging at a wide range of distances from less than one meter to 200 meters or more with centimeter resolution.
In FMCW, unlike ToF, rather than sending a short pulse of light out and looking for its return, the laser transmits a continuous sweep of wavelengths of light. As in ToF, the goal is to look for the return signal, but with FMCW detection, the sensor retains a portion of the transmitted signal, and this retained piece is combined with the returned signal, which produces a multiplicative effect that boosts the signal and provides much greater sensitivity, which is critical to the distance that a LiDAR sensor can “see.”
Autonomous vehicle designers and other software stack developers that are creating “drivers” for these vehicles are looking for LiDAR sensors capable of seeing objects 200 meters away with reflectivity as low as 10 percent, which is roughly the reflectivity of a car tire. When traveling at highway speeds, 200 meters gives an autonomous vehicle only about six seconds to identify the object and take appropriate action. Traditional ToF systems have difficulty in meeting that 200-meter range requirement.
Most ToF LiDARs operate somewhere between 850 and 905 nm. There is a limit to the amount of laser power that a LiDAR sensor can emit at that wavelength and still remain “eye safe,” i.e., it doesn’t run the risk of blinding someone. A new class of ToF lasers operates at 1,550 nm, which allows the use of higher power while remaining eye safe, but the sources and all the associated components are far more expensive than the 850 to 905 nm lasers, so they can’t meet industry cost targets of approximately $250 for the complete LiDAR sensor system.
Insight’s FMCW LiDAR, with its higher sensitivity, puts all of the optical components onto Photonic Integrated Circuits (PICs), enabling low volume cost. In addition, Insight’s swept-source and dynamic software controls — which were originally developed for medical imaging — offer high resolution, so even a small object like a child can be detected and identified more than two football fields away.
A variety of technological challenges must be overcome before autonomous vehicles can become an everyday reality on our roadways. Currently, the autonomous vehicles being road-tested employ ToF LiDAR sensors. Insight’s FMCW LiDARs are now in the early stages of being evaluated for these applications. The autonomous passenger vehicles of the future will likely combine ToF LiDAR (for shorter-range imaging) and FMCW LiDAR (for longer-range imaging).
For Free Info Visit Here