[Dprglist] Fwd: PAA5100JE Near Optical Flow SPI Breakout is now available at Pimoroni!

[Murray Altheim]

Hi Karim,

I'll try to get the post, which is about half ready, posted ASAP, though I've
got a lot going on right now at home and work.

I think it's definitely worth a look. My initial play with it was using the
existing PMW3901 Python library, as suggested by the Pimoroni techs, and I'd
tested it out over nine different surfaces and with that library I found the
image complexity was clearly related to the measured velocity, but Pimoroni
updated their library, and while the update doesn't entirely eliminate that
issue, it's much less significant than before. I don't think you'll be able
to use the sensor solely to measure distance over significantly varied
surfaces with accuracy, but over a relatively constant surface I think it's
pretty good, within a few percent.

It outputs a simple stream of x,y values when moving, clearly on an interrupt
basis as that stream stops when the movement stops. I measured the 'x'
dimension (oriented as it was) over one meter and so long as I stayed on the
kind of surface that I calibrated it for the measurements were quite accurate.
An interesting notion might be to use the other axis with a PID controller to
travel in a straight line, as while dragging my little robot sled over the
test surface I'd get small 'y' values that could be used for course correction.

With the new Python library I think the minimum to maximum of the nine
surfaces I tested on (tile, wood floor, lawn, concrete, rug, etc.) varied by
something like 10-15%, but on any given surface the ten runs of my test
varied by only a few percent max, often less than 1%. Repeatable results.

It'd be possible to calibrate the sensor over different surfaces and switch
calibrations based on the surface if it were possible to detect the surface
type, possibly using a camera or spectrometer sensor like the Pimoroni
AS7262 that could detect color (which actually works pretty well, e.g.,
it's easy in my house to know the difference between my tile, wood and
carpet flooring by color). I've got a spreadsheet of results that will be
part of the review/blog post.

At £18.90 (US$26.75 / NZ$37) it's not cheap but I think worth a look, yes.
The only downside really, which was entirely predictable, is that its
operating range is 10-35mm, suitable for hanging underneath a small robot,
but not so much under say, a Mars rover, whose bottom is likely higher than
35mm. In that case the PWM3901 would be more suitable, but not necessarily
as accurate,

Cheers,

Murray

On 9/06/21 8:52 am, Karim Virani wrote:
> How about at least a teaser? Worth a look?
> 
> On Tue, Jun 8, 2021 at 3:49 PM Murray Altheim via DPRGlist <dprglist at lists.dprg.org <mailto:dprglist at lists.dprg.org>> wrote:
> 
>     Hi,
> 
>     FYI, Pimoroni has finally announced stock availability on their new PAA5100JE
>     Near Optical Flow Sensor, which I'll be posting a review on hopefully soon.
> 
>         PAA5100JE Near Optical Flow SPI Breakout
>         Detect motion of close-up surfaces with this Breakout Garden compatible
>         optical navigation breakout - great for floor tracking for ground dwelling robots!
>     https://shop.pimoroni.com/products/paa5100je-optical-tracking-spi-breakout <https://shop.pimoroni.com/products/paa5100je-optical-tracking-spi-breakout>

...........................................................................
Murray Altheim <murray18 at altheim dot com>                       = =  ===
http://www.altheim.com/murray/                                     ===  ===
                                                                    = =  ===
     In the evening
     The rice leaves in the garden
     Rustle in the autumn wind
     That blows through my reed hut.
            -- Minamoto no Tsunenobu