0

I had an idea for a LiDAR scanner but am having trouble locking in the feasibility of it.

First of all - I know stereo matching would be way more efficient - this is more of a thought experiment at this stage.

Usually LiDAR scanners are directed by either single or dual axis robotics, generating either a single axis scanline or a full 3D model.

What I am wondering is the feasibility of creating a LiDAR rig that scans within a 2D window, much in the same way a camera takes in light information and eventually ends up as a 2D pixel map. The purpose of this data being for real-time recreation of an environment, requiring 30+ frames generated per second.

The data collection area would be similar to that of a stereo matching device - a field of view around the direction the unit is pointing at would get scanned.

I imagine the major data acquisition bottle neck would be the rotors the sensor is mounted on - how quickly can they traverse the acquisition window. I believe there are LiDAR units with a high enough point collection rate for the task, but I am unsure about the robotics involved.

As an example, let's say that the task was to collect point data in a 60° x 40° window in front of the rig, with a point density of 100 x 70 and at 30 ticks per second. 7,000 points must be collected per frame to populate the region, multiplied by 30 frames per second = 210,000 points per second. Not a huge ask for quite a few modern sensors. Where I'm getting stumped is the best way to construct a robotics rig that will direct the sensor to the angles it needs to collect the data at, if it is even possible to move a sensor that fast.

FacePuncher7
  • 1
  • 1
    Welcome to Robotics FacePuncher7, but I'm afraid that Unbounded Design Questions are off-topic because there are many ways to solve any given design problem. We prefer practical, answerable questions based on actual problems that you face, so questions which ask for a list of approaches or a subjective recommendation on a method (for how to build something, how to accomplish something, what something is capable of, etc.) are off-topic. Please take a look at [ask] & [about] for more information on how stack exchange works. – Ben Mar 14 '21 at 22:46
  • This is all hypothetical, so there's nothing firm to give feedback on here. Can a sensor do it? Most laser scanners use a spinning head to turn basically a mirror inside the scanner to choose the "firing angle" for the laser(s). You could put a prism with 72 facets, one for each vertical scan line, inside a spinning head; each facet would scan a new vertical line. You could mount a mirror to acoustic drivers tuned to the horizontal and vertical scan rates. You could have 72 lasers firing through the same head (Velodyne already does 64). You could do solid-state lidar, or fiber optic, etc. – Chuck Mar 16 '21 at 12:36
  • The point is, there are lots of ways to solve a problem. It's up to you to pick a solution, but the solution that's best for you is specific to your problem and your needs and unlikely to help future visitors to the site. This is why the unbounded design questions are off-topic. Most of the robotics equipment will not be "commercial off-the-shelf" (COTS) solutions, but most of the lidar equipment is not COTS, either. Eye-safe lasers powerful enough to have meaningful range, not in the visible spectrum, laser detection sensors, high-speed time-of-flight chips, etc. – Chuck Mar 16 '21 at 12:43

0 Answers0