Re: [mowbot] cutters

Mike Ross (mross nospam at outland.jsc.nasa.gov)
Tue, 8 Oct 1996 12:24:44 -0500 (CDT)

On Tue, 8 Oct 1996 robin nospam at acm.org wrote:

> I don't think this would work exactly as described. The cylinder doesn't
> have any useful inertia to cut against unless it is already rotating.

You're thinking angular momentum; the inertia of the reel is the same
regardless of velocity. I'm thinking of having a many-bladed reel, though
12 blades is probably the limit. It should have enough inertia to absorb
the small impulses caused by the grass cutting, and yet be hand
back-drivable.

>
> Remember also that such a cutter cuts only the one little patch where the
> rotating and bottom blades intersect. The cutting patch moves
> sideways as the spiral of the moving blade rotates. To make it cut a nice
> even swathe as the mower moves forwards, the spiral blade has to rotate
> quickly enough to make the ``raster scan'' of cutting patches overlap.
> So the rotating cylinder must be goinDg pretty fast. ...

Good point, I was assuming this thing is crawling along, such that a 10
rpm reel speed would be enough to 'raster-scan' the grass.

>
> We might be able to design something based on the cylinder principle that
> cuts the grass ``sideways'', so the vibrating cutter isn't working against
> the rotation of the cylinder (e.g. with a zig-zag edge on the bottom blade
> which oscillates from side to side), but the construction of such a design
> is likely to be a big engineering challenge that most of us aren't up to.
>

Well, from my perspective, it is the CENTRAL issue. All the rest of this
stuff is, while not trivial, of hobbyist level now. 68HC11's make
embedded processors easily available, algorithms exist for mapping,
battery technology is *adequate*, sensors are cheap now. The only thing
keeping this dude from the market is the fear of venture capitalists that
a lawsuit will take them down. This thing HAS to be able to run over a
baby (or RELIABLY avoid it) without making the baby even pout. Other
people have traveled this road before. A guy I saw at a robotics
conference was trying to drum up money for a line-of-sight teleoperated
lawnmower, which was a typical rotating blade on a cybermotion style
platform. He had a prototype, a promo video, and brochures, yet he told
me he had not had one nibble in two years of searching. All because of
liability *perceptions* on the part of venture capitalists. It will take
an obviously benign mechanism to offset those fears. That's where I feel
this discussion is handy: somewhere out there, the basic mechanism has
occurred to someone and perhaps he or she would like to share it and
remove this last real barrier to a cheap autonomous lawnmower.

Of course, I like the idea *I* proposed :-)

-mike

* Michael L. Ross/C33 | Lockheed Engineering & Sciences Co.****
* Robotics Department | 2400 Nasa Rd. 1, Houston, TX 77058*****
*(713)333-7094 voice,(713)333-6071 fax*#include <disclaimer.h>*
* Finger me for my public key. mross nospam at outland.jsc.nasa.gov ****
* If you weren't a socialist in your twenties, you had no heart.
* If you weren't a conservative by your thirties, you have no brain.