[Dprglist] New lidar technology

[Doug Paradis]

Paul,
     Interesting article. Claims large scale sampling in 2025. Do your
sources have any info on its capabilities?

     On the other end of the scale, DFRobot has a new product with 400cm
range available,https://www.dfrobot.com/product-2727.html. One flaw is a
single I2C address, making multiple sensor applications more difficult.
Regards,
Doug

On Tue, Aug 22, 2023 at 4:53 AM Paul Bouchier via DPRGlist <
dprglist at lists.dprg.org> wrote:

> OK - maybe not really new technology so much as combining known techniques
> in a new way and applying them to lidar...
>
> DPRG folks,
>
> I saw an article in IEEE Spectrum here:
> https://spectrum.ieee.org/lidar-on-a-chip
> describing lidar-on-a-chip for automotive applications. It's being called
> 4D lidar, because in addition to providing range across a 2D field of view
> (thus providing a 3D point-cloud) it also provides velocity toward or away
> from the sensor for detected objects (hence the 4D). And it targets a $500
> pricepoint.
>
> The techniques that are new to lidar (AFIK), but have been used in radar
> for a long time, are:
> 1. FMCW - Freqency Modulated Continuous Wave lidar. They change the
> emitted light frequency (called chirping in radar) and compare the emitted
> frequency with the reflected frequency to determine range and speed.
> 2. Phased array emitters and receivers produce an electronically steerable
> beam (called synthetic aperture in radar) to produce an electronically
> direction-controlled beam that can be steered to scan the scene.
>
> FMCW
> This involves "chirping" or changing the light frequency in a repeating
> fashion, via a triangular frequency modulation of the laser. When the
> reflected light is mixed with a local copy of the currently-emitted light,
> the frequency difference between the two is proportional to the range
> (because the local copy of the emitted light will have changed frequency
> between the time it was emitted and the corresponding reflection is
> received). Heterodyning provides the sum and difference frequencies which
> can be measured to provide range. Furthermore, if the target is moving
> toward or away from the sensor, the reflected frequency is modified from
> what a stationary object would have reflected due to the doppler effect.
> Comparing the reflection from a rising chirp with a falling chirp enables
> separating the doppler frequency shift from the range-related frequency
> difference as described (for radar) here:
> https://www.byte-lab.com/fmcw-and-the-doppler-effects/
> Looking at multiple heterodyne frequencies enables separating objects at
> different ranges that both fall within the detector beam width (imagine two
> distant cars on a road, one closer than the other, one traveling away from
> you and the other toward you).
>
> Phased array
> A phased array antenna introduces a phase difference between adjacent
> closely-spaced antennas so as to control the direction of the beam, which
> can be done in 1D (linear) or 2D. The net of this is to produce an
> electronically steered laser beam that can be moved at megahertz-level
> rates. The same principles apply to the receiver. The antennas are about a
> wavelength apart, which means for infrared the lasers are about 1.5
> micrometers apart and highly coplanar and there are thousands of them.
> Obviously this can only be done on a chip. There's a good graphic in the
> article illustrating how beam steering via multiple phased antennas works.
> Doing this with lidar involves introducing a variable delay in the light
> emitted from each laser. They appear to do this with an electro-optical
> element in front of the laser, where an applied voltage changes the speed
> of light propagation and thus the phase of the light emitted from a laser
> relative to the one next to it.
>
> I would expect initial versions of this kind of chip to initially be
> limited to auto manufacturers (specs and devices), but maybe over time as
> they hone the techniques, they may become more widely used and available.
> These sensors are qualitatively different than previous lidar sensors, and
> seem to take lidar to the next level by far exceeding the capabilities of
> the old "spinning coffee can" Velodyne ($75k) sensors.
>
> Whew! Sounds like a good RBNV discussion topic.
>
> Paul
>
> _______________________________________________
> DPRGlist mailing list
> DPRGlist at lists.dprg.org
> http://lists.dprg.org/listinfo.cgi/dprglist-dprg.org
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.dprg.org/pipermail/dprglist-dprg.org/attachments/20230822/0f0c8f1b/attachment.htm>