From: Brent Roman [brent@mbari.org]
Sent: Monday, May 23, 2005 11:18 PM
To: Cline, Danelle
Cc: Radochonski, Wayne; Meese, Timothy; Jensen, Scott
Subject: Re: Benthic Rover controller and OS discussion tomorrow 11 am
Surface Commons ?

Hi Danelle,

I'm helping Chris and Roman with protocol development tomorrow.  I'll try to get them to a state where I can break away by 11AM.  However, Scott or I may need to stay in the lab.

We'll bring a couple vendors' PC/104 ARM CPUs for show-and-tell :-)

I've exported a .pdf version of our review from last December to my tempbox:

//tempbox/brent/softdesign.pdf

It has a few slides you might find interesting.

- brent


Cline, Danelle wrote:

>Hi Wayne, Tim, Scott and Brent,
>
>I'd like to talk with you about controller and operating system options 
>for the Benthic Rover project.  Could you meet tomorrow at 11 am in the 
>Surface Commons to discuss this ? Please send me a quick email to let 
>me know if you can meet, or suggest another time if that doesn't work 
>for you.
>
>The Benthic Rover project is a ~10 year old WHOI project our new 
>scientist Ken Smith has brought with him. What MBARI is doing is 
>building a second generation Benthic Rover with a modern controller and 
>COTS instruments where possible.
>
>What I want to talk about is:
>
>Could we reuse any of your existing controller boards ?
>
>What are your OS suggestions ?  I'm also particularly interested in 
>hearing what kinds of cross-development tools you've had success with, 
>and how power management is done with the ARM processors
>
>The Benthic Rover needs a design that supports: storage (e.g. compact 
>flash), a USB port, RS485 and RS232 serial ports for COTS instrument, 
>and both brushed and brushless stepper motor control.
>
>The Benthic Rover requires power management for the central controller 
>and all peripheral devices. It's deployed for 6 months on battery power.
>It function is basically to (1) crawl to a new site taking a few 
>pictures along the transit, then (2) when at the site, takes a slew of 
>measurements (typically from serial instruments and a few frame grabs), 
>then (3) go back to sleep until the next sample. The two continuously 
>powered components are the central controller and a stir motor in the 
>benthic respirometer. The only hard real-time requirement is for motion 
>control, which is probably best done in firmware.
>
>Thanks for your help.
>
>Danelle
>====================================== 
>Danelle E. Cline, Software Engineer R&D Division        
>Monterey Bay Aquarium Research Institute http://www.mbari.org
>(831) 775-1947 voice
>(831) 775-1652 fax
>======================================
>
>  
>


--
 Brent Roman
 Software Engineer                 Tel: 831 775 1808
 425 Clinton St., Santa Cruz,      California, 95062
 mailto:brent@mbari.org  http://www.mbari.org/~brent