8 comments

[ 3.5 ms ] story [ 27.2 ms ] thread
Cheap laser rangefinders are absolutely an enabling technology for the next generation of robotics. I sincerely hope they can deliver.
I have been wanting one of these for years! Last time I looked into building my own robot, the lack of a low-cost LIDAR is the thing that held me back. Last time I looked, it was like $10k for similar functionality. This is a total game-changer. This will also be my first Kickstarter to back.
Go buy any vacuum cleaner from Neato, cost: $300-$500. Give it a LIDAR-ectomy and also save the nice battery pack. Toss the vacuum parts. Now you have a nice LIDAR with a well-known interface protocol (thy friend is Google). This is the cheapest good LIDAR around. 360 data points, 5Hz scanning, 1 degree resolution, 5m indoor range, 2+ meters outdoors if you provide a sunshade. Or, if you like the Neato base, use the USB protocol and Neato ROS node to drive the unmodified vacuum around.

The Neato LIDAR is a structured-light Laser Distance Scanner (LDS), accuracy about 1-2cm depending on distance. Resolution 1mm. Minimum sensing distance about 20cm, IIRC.

That's definitely mentioned in the article. :) Downside is the reduced range (see previous comment). I wish Neato would make their LRF available separately for purchase! I even wrote about it back when it was under development in 2009!

http://www.hizook.com/blog/2009/12/20/ultra-low-cost-laser-r...

Yes, well, a lot of us have told Neato the same thing. They are not interested in pursuing the OEM market. I can see their point.
FWIW I know one of the guys behind the Scanse LIDAR. Pretty sharp guy. I saw a demo a few days ago at the Homebrew Robotics Club. Pretty decent specs, although the minimum distance using the current software is a little on the long side -- at least 1/2 meter IIRC. The max distance is huge. They say this is a true time-of-flight LIDAR. IMHO this is an interesting unit and I am confident enough that I put in a Kickstart pledge. One interesting point is that the RPM is independent of the point update rate, so you can trade off resolution agains full-360-scan update rate. I have enough confidence in the founders and what I saw to believe I will get a decent unit, perhaps with minimal, reasonable, schedule slip (this stuff is hard).
(Author here) RPLidar is great for home use... but the 6m range is severely limiting outdoors and in larger indoor environments such as warehouses, retail stores, etc. Anecdotally, you really need at least a 15-20m range to effectively map and localize in such environments -- which is the sweet spot for the Scanse.