AUV docking meeting
wed 12/22/04 1000

Zhang,Bellingham,McBride,Erickson, Thomas,Thompson
Godin, McEwen, hoover

Science applications, cost of building AUV


# Mission 2: Pajaro River Plume (paull,ryan)
Elkhorn vs Pajaro

pajaro safer
elkhorn: multiple processes, mixing, power plant

pajaro preferred; easier to distinguish input contributions
plume dies down after a couple of weeks.
pajaro has pretty good RT monitor (USGS monitor station, event detection can 
be done on shore)

is there a hyperpycnal flows? may flow down into canyon.
spatial scale:5 km x 100m x 10s of meters (lxwxd)

trigger mechanism: shore signal (river discharge), in-situ can use velocity 
sensor on vehicle or dock.
frequency: usually only happens in December

Q how is dock instrumented?

Why is vehicle needed (can be done as well with surface ships etc)

# Mission 2: Turbidity Flow /gravity (paull)
lower canyon dynamics, "landslide" in the canyon
often triggered in winter. flow goes down narrow canyon, spreads into basin.
proposal: put docking station in basin (1850m depth)
some mooring data has been collected at single point along canyon.
AUV measurements would give better view of process.

1850m depth in basin (beyond limit of Ventana's depth)
scale: transects across basin 1.5 km diameter, 200m above seafloor
duration: unknown
measurements: optical, ctd, ADCP
triggers: backscatter, current

# 
do we just want a dock? instruments? MOOS BIN + dock?
needs comms path for launch code
dock must be a pass thru for data
BIN not deployed yet.
Data logging on dock?
1.5 hour transit, 1.5-2 hours to descend: 0 bottom time.
no current, no wind
Power concepts: 
- first time just use primaries on vehicle (no recharge), just multiple demo 
vehicle responses in situ.

could dock closer, and transit vehicle
distance to site: near M2?
docking farther: transit power, response time

event frequency: happens in Nov-Mar, 3-4 x per yr

currents at various depths...

adcp, dvl, 100m -- should be more like 200 m
(should try to keep dvl away from propulsion)

# 
Elkhorn ruled out, pajaro more attractive, though close to shore, logistically
 convenient; what is science value?
transponder array vs surface gps fix -- difficult on bad days in shallow water?

turbitity flow: not a beginner experiment. complex mission, 
some careful trades between docking location

intermediate site for docking testing? early docking experiment w/ sci results
w/o a lot of testing. (ref ice sheet experiments)

Leslie Rosenfeld (NPS): pumping processes in canyon generated by internal tides
strong benthic currents generated by internal tides. based on mooring currrent
meas.

duty cycle, duration of deployment (seems like you'd get more info if AUV is
 relatively busy)

ground truthing: are AUV only meas's? are there other independent meas's?
(did vehicle error, was there aliasing? ...)

dock protection vs strategic meas location: hopefully not in the middle
of a violent process.

midwater experiments might be good, safe environment -- question is what is 
events, just video data is no good. 

# response time requirement needs to be defined.
# first dock location vs final deployment locations. need good access.
# No docking mission in 2005: build docking system and do experiments
# make first demo of docking capability; show that it won't eat too much 
# resources, not too expensive. Get sci interested in docking-based missions
# Done some development of docking vehicle concepts.


################
AUV build cost/time discussion

bellingham, erickson, thomas, mcewan, hoover

replication cost
multivehicle support
expensive vs inexpensive economies of scale, support ships, dock cost

how to quantify vehicle cost?
jim's rule: when building a system from many small components
take parts x 4 (small company w/ lo overhead)

if buying large integrated parts, don't mult x4

if cheap parts are available, is it more expensive to build or buy?
depends on relative labor cost and assembly complexity.

also must consider how we want to spend our time (core competency)
e.g. fairings

design for manufacture is important: must design for appropriate scale.
tooling investment is $$ for larger scales.

we have materials cost, but would be good to do labor flow chart, analysis

no good rule of thumb for build vs buy based on material cost

we have to think in terms of having someone else reproduce it.

design for mfgability early is important.

integration, calibration, test time is part of vehicle build cost.

fairing design is a place where  improvements could be made.

must consider total lifetime of vehicle costs: mfg, integ/test, operation/maint

depot maintenance vs field maintenance

revelation: pay attn to parts cost, but real cost
 (and opportunity for change) is in labor.