[Dprglist] [Dprg_roborama_committee] DYI 360 degee spinning LiDAR

David Anderson davida at smu.edu
Sun Dec 20 22:48:03 PST 2020 

Chris,

Cool.   I've done some similar things with sonar readings while 
rotating, though not nearly as sophisticated.   I had a couple of thoughts.

As per dewarp_scan.py:

# what's left to do is to remove the artifact that appears in this 
dewarped scan.
# when the robot is twisting right, then the artifact is on the left and 
vice a versa
# root cause is unknown
# as a workaround, just filter out (i.e. invalidate by setting to inf) 
readings in a certain angle range input

I wonder if this is an interaction between the spinning mass of the 
laser and the spinning mass of the robot?  Either as an additive 
process, like if the robot is rotating the same direction as the lidar 
or not,  or as some sort of actual mechanical gyroscopic-type coupling 
between the two?

My outdoor robot moves along at a brisk walking pace, about 5~6 
feet/second, certainly not turtlesque.   It's array of 4 sonar can see 
out about 20 feet across a FOV  of about 60 degrees, and reads all 4 
sensors every 120 ms, so maybe 8 Hz.   I've found as a practical matter 
that update rate is plenty sufficient for outdoor navigating in some 
pretty intense environments, with lots of obstacles and treachery.

So it's not clear to me why you find that a higher update rate is 
required, particularly given the increase in data density that you 
already get with the lidar?

best,

David











On 12/20/20 5:02 PM, Chris N via DPRGlist wrote:
>
> My 2 cents:
>
> Definitely an impressive design, i.e. from an electro-mechanical point 
> of view, but given that the cost of the RPLidar A1 (and its clones) is 
> <$100, I’m just not sure its worth it.  Maybe it has a form-factor 
> advantage in some situations?
>
> Since it has 2 sensors, the effective rotation rate is actually 
> equivalent to 120RPM, but that is still too slow!
>
> I find that even the 300RPM of the RPLidar A1 is too slow for a robot 
> that moves faster than a turtle – at least in my experience with the 
> navigation stack in ROS (e.g. Gmapping or hector slam).
>
> I have cranked it up to almost 600RPM (10Hz), but at that rate, my 
> Pololu Romi base was shaking quite a bit…..I suppose on a larger / 
> heavier base this would be fine and I should probably try that at some 
> point.
>
> The problem with 300RPM is that the laser scan gets significantly 
> warped as a result of robot motion, especially when the robot is turning.
>
> I might have mentioned it already to this group, but I have actually 
> spent significant amounts of energy on a lidar data de-warping logic 
> earlier this year.   Got it to work, but it didn’t seem to improve the 
> SLAM performance much – there are probably navigation stack parameters 
> that need to be tuned properly…….That’s just one of the many 
> complexities in ROS……
>
> My code for the Lidar de-warping is here: Experiments with de-warping 
> lidar scans <https://github.com/nettercm/lidar_dewarping>
>
> Discussion on answers.ros.org: How to deal with slow RPLidar scan rate 
> on fast moving robot? 
> <https://answers.ros.org/question/242658/how-to-deal-with-slow-rplidar-scan-rate-on-fast-moving-robot/>
>
> The code is in the form of a ROS node, but the core logic can be used 
> outside of ROS of course.  Anyway, if somebody is doing Lidar and is 
> finding the speed of the RPLidar to be an issue (due to warping), let 
> me know and I can provide more details.
>
> It was kind of fun (but exhausting) applying some math to create this 
> de-warping logic, but if at some point I decide to get serious again 
> about ROS and Lidar, maybe I will just pony up the $600 for a RPLIdar 
> A3:   25 meter range,  16K samples /second,  1200RPM (20Hz). 
>   Monitoring eBay for a good deal…..
>
> Chris.
>
> *From: *Doug Paradis via DPRGlist <mailto:dprglist at lists.dprg.org>
> *Sent: *Sunday, December 20, 2020 12:49 PM
> *To: *Ron Grant <mailto:deltagraph at aol.com>
> *Cc: *dprg_roborama_committee at lists.dprg.org 
> <mailto:dprg_roborama_committee at lists.dprg.org>; 
> dprglist at lists.dprg.org <mailto:dprglist at lists.dprg.org>
> *Subject: *Re: [Dprglist] [Dprg_roborama_committee] DYI 360 degee 
> spinning LiDAR
>
> Ron,
>
>      I agree about the rotation rate. I am also not a big fan of the 
> VL5xL0X sensors, but I haven't tried the VL53L0x. I do think the 
> project shows the issues that need to be addressed. If I were just 
> starting out with nothing, I would prefer to just buy a RPLidar A1M8 
> from Robotshop.or DFRobot.  It has a scan rate between 1 and 10 Hz 
> with 5.5 Hz used to get 1 reading per degree.
>
>     For indoors, this would be a good unit to have in your new 
> simulator. For outdoors (and also indoors) you might want to allow for 
> the Benewake line of sensors and the LiDAR-Lite. These are not 
> scanning LiDARs, but have larger range and are usable outdoors.
>
> Regards,
>
> Doug P.
>
> On Sun, Dec 20, 2020 at 9:49 AM deltagraph--- via 
> Dprg_roborama_committee <dprg_roborama_committee at lists.dprg.org 
> <mailto:dprg_roborama_committee at lists.dprg.org>> wrote:
>
>     Looks cool, but kinda slow rotation rate at 60 RPM vs about what
>     5X to 10X that speed in some of the robot vac cleaners.
>
>     Of course not bad to stop for a second, scan then move on, OR at
>     least be aware of your position vs scan vector.
>
>     I am working on a new simulator rendition with LiDAR, looks
>     promising and fun to play with.
>
>     I can see useful to include variable scan rate to simulate range
>     of available sensors low-cost units to high-speed high dollar ...
>
>     Ron
>
>     -----Original Message-----
>     From: Doug Paradis via DPRGlist via Dprg_roborama_committee
>     <dprg_roborama_committee at lists.dprg.org
>     <mailto:dprg_roborama_committee at lists.dprg.org>>
>     To: DPRG <dprglist at lists.dprg.org <mailto:dprglist at lists.dprg.org>>
>     Cc: Doug Paradis via DPRGlist <dprglist at lists.dprg.org
>     <mailto:dprglist at lists.dprg.org>>
>     Sent: Sun, Dec 20, 2020 9:20 am
>     Subject: [Dprg_roborama_committee] [Dprglist] DYI 360 degee
>     spinning LiDAR
>
>     Interesting Instructable:
>
>     https://www.instructables.com/Project-Lighthouse-360-Mini-Arduino-LiDAR/
>
>     Regards,
>
>     Doug P.
>
>     _______________________________________________
>     DPRGlist mailing list
>     DPRGlist at lists.dprg.org <mailto:DPRGlist at lists.dprg.org>
>     http://lists.dprg.org/listinfo.cgi/dprglist-dprg.org
>     _______________________________________________
>     Dprg_roborama_committee mailing list
>     Dprg_roborama_committee at lists.dprg.org
>     <mailto:Dprg_roborama_committee at lists.dprg.org>
>     http://lists.dprg.org/listinfo.cgi/dprg_roborama_committee-dprg.org
>
> _______________________________________________
> Dprg_roborama_committee mailing list
> Dprg_roborama_committee at lists.dprg.org 
> <mailto:Dprg_roborama_committee at lists.dprg.org>
> http://lists.dprg.org/listinfo.cgi/dprg_roborama_committee-dprg.org
>
>
> _______________________________________________
> 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/20201221/3422d72e/attachment-0001.html>

More information about the DPRGlist mailing list