Microships Wordplay & Songline
(and the kayaks, Bubba & Stella)
Project Management and Task List
Last Updated:  Feb 15, 2006
Steven K. Roberts
Nomadic Research Labs

If you are involved with the Microship project, or just want to help in some way, this is the place to look...

I’ve partitioned this into broad categories to help keep it manageable, and each topic opens with a little explanation of what we’re trying to accomplish. Below that, it’s broken down into Clearly Defined Tasks, or CDTs – the finest level of project granularity that can be expressed (or that I recognize at the moment). While this sometimes appears overwhelming in sheer volume, it offers the advantage of well-stated jobs rather than vague objectives. Of course, some projects are not yet clearly understood.  Tasks that have been completed are lined out; those that are being deferred until later are light gray to render the list a bit less overwhelming. 

If you want to keep an eye on what has been done, please check the daily update page for current activities and photos. 

Kayak Refinements

A 26-mile solo paddle last fall with full touring gear gave me the opportunity to test Bubba (the Aire "Sea Tiger" inflatable  kayak that will serve as towable shore-support pod), and its integrated systems including APRS tracker, solar charge management, navigation, and other tools.  Mostly, it worked beautifully, but I returned with a TO-DO list (of course).  The current plan is to do a short test (3-4 days) over to Hood Canal around the end of June, then launch southbound from Port Hardy in early August.  Gray items do not really have to be done by August, but are still on the list.

Service stands for Stella
Replace Alinco DJ-5T with waterproof Yaesu VX-7R
Investigate writing "commander-like" VX-7R interface software in Squeak on the PDA
Cellular phone is, alas, a necessity
Acqure Dell Axim X3 as cheap portable Squeak platform
PDA in Otterbox instead of laptop; 802.11 to server on each substrate
Docking bay in aft bin
iPod for music library; FM xmtr interface
Cartography for Garmin GPSMAP-76S to augment basemap.  (Need Windoze box to install)
Better food planning, including recipe database and shopping-list generator.  Was too haphazard.
Improve clothes packaging; this was sloppy.  Eliminate cotton.
Checkout system for items removed for use when not traveling, to simplify next launch prep
Fix MSR fuel pump knob
Improve sleep system (knee pillow for lumbar pain)
Acquire new sleeping bag
Boeshield on Hustler bayonet antenna mount; showing corrosion
Install old Amp-Hour meter guts in power/tracker box
Replace current AC charger
Establish packet accounts and get through learning curve
New Werner Camano bent-shaft paddles
Acquire Kokatat OutFIT Tour PFD and keep SOSpenders inflatable for Wordplay/mothership
Better trash management
Move rudder pedal tracks 2-3" forward (same cable length; sliders are bending and raising friction)
Flares (get jammed Olin flare gun replaced)
Replace man-hour extension facility with lidded model for tent use as well
Replace final cheesy Seafit shock-cord assembly on gray box with something reliable
Deal with corroded zippers
Fiberglass gear enclosures to replace dry bags, straps, and tarps:
Build up in situ foam-core structure with .5" Divinycell and hot glue, shaped to hull
Order suitable quantity of 10-oz cloth and epoxy
Spec D-Ring fixtures for end handles (handle points, lid lashing, and  tie-downs)
Lid gasket design,  overlapping engagement, distal hinges
Southco latches for lid clamp-downs, including on long edges at internal bulkhead (or straps/buckles)
Vented solar mounting fixture (forward) and Paddleboy cart (aft)
Lid internals: flat stowage for charts/books
Possible integrated insulated space in forward unit as cooler
Glass/goo structure and test
External fairing
Paint red to match Sea Tiger

Do similar bin(s) for Stella
Add better hauling handles to Bubba (like Stella's); current ones are painful
Build lash-on rudder platform for Stella, and add Feathercraft retrofit
Towing test behind Wordplay

Nautical and Mechanical


The 2001 mini-expedition was useful in many ways, one of which was to reveal an urgent list of leak fixes written during a grim 3-day rainy layover in Everett Harbor. Waking up every couple of hours to sponge out the bilge is no fun. Here’s what needs to be done, though the boom tent and the use of Bubba to camp on shore may make most of these less urgent for now:

Seal pinhole leak(s) at daggerboard trunk with low-viscosity adhesive
Sealant around bilge pump exit into aka nest
Slow drip via sternlight wires
Furler, landing gear, and mainsheet line entries to cowling....little hoods?
Aka notches (see dodger jobs)
Replace current cowling solar mounts with button-heads, gaskets, and goo; add pads for windshield
Epoxy-fill wind-sensor mounting holes while goo available for solar panel
Ackerman tube foam-fill
Arch tube foam-fill after all cabling done
HF antenna mount

Random Jobs and Gear Issues

This is just a catch-all category for little stuff.
More freshwater cleanup – still salty in lots of places
Unroll rig and give it a bath
Sand gray uglies below waterline and wax hull (later)
Shorten seat yokes, alodine, and pad them... head bangers when sleeping
UHMW tape at aka joints... creaks and clunks
Better coding of aka pins
Change WN0536NL numbers; add new registration
Better ama pads on arch for road mode (neoprene is fine)
Place McMaster-Carr and West Marine orders
Deploy new tent to get familiar with it, and mate with red dry bag
Test camp gear overnight
Complete gear list and aggregate into dry bags
Pack needed charts and reference material
Move default rudder-pedal tracks on Bubba to solo position
Other Bubba setup, gear fixturing, etc.
Confirm FCC licensing; upgrade as needed to allow Canadian operation
Disinfect fresh-water tank

Fabric Projects

This includes the dodger that covers the entire cockpit area, as well as related projects.
Re-do side windows with single zippers. Bottom flaps indep of solar.
Aka notches need good closure for airflow and water (in progress)
Blackout windows
Boom tent for an extra layer of protection - add wing straps after test fitting. Aft addition needed?
Sew Tyvek into 7.5-foot square footprint for Prophet tent

Cockpit and Gear Issues in General

Mostly, this is all about making life more pleasant on-board: minimizing clutter, disambiguating stowage, increasing comfort.
Establish gear shelving area in lab
Stowage (fwd hatch?) for Ackerman handle; it wears socket when left in place on water
Better water bottle cage than floppy waterboy thing
Improve "man-hour extension facility" mounting behind seat
Flute stowage tube (later)
Eliminate all HT’s and standalone un-integrated junk (later)
Rear-view mirror essential; commercial boat ones too clunky
Consider manual windshield wiper, if not too big
Improve sleeping system: softer Therm-a-rest, inflatable knee pillows
Check windshield opening interference from cowling solar panel
Charger stuff hard to use, build in all functionality (later)

Pedal Drive Unit and Crankset

The Spinfin pedal-drive unit worked very well during the test sail, though the temporary test crankset was fiddly and hard to remove every night for sleeping (Tran-torque couplers are great for some things, but not this). Besides, it was made of Crome-Moly steel, and is rusty; it was only intended to be a test unit.
Inspect and improve drive unit gasket
Reassemble with 0w30 oil and gasket goo. Use lots of oil.
Start designing new pedals for songline cranks
Assess original parts for interface with existing bearings (later)
Find shaft coupler for removability (later)
Stowage clip somewhere so it’s not on afterdeck... possibly under deck?


The big refinements to the WindRider rig were completed before the last trip, and generally work well. Still, the furler needs a little improvement. There are some tweaks and a few new projects...
Change furler to 1/4” line (acquired, now rig it and test for adequate spool depth)
Add strap eye in outer quarter of boom to capture main (flops about)
Remove masking tape from mast above furler drum
Furler line:  change to a better cleat design in same location
Change mainsheet to 5/16; the current 3/8 is a bit much.  Acquire 50' of it.
Shore tie line – 400’ polypro reel in fwd hatch; establish deployment protocol
Flat fender for softer dock impact on each side, captive under solar array and hung from rail
Hank of line and two triple blocks for boat handling, hoist, etc (secondary use of shore tie line)
Towing bridle for Bubba:
    Fabricate bails on traveler ends with SS bar stock
    Bridle, X-configuration, polyprow tow line to Lewmar nabs and bails
Frapping on all line ends (thanks, Marcy!)
Trim seized line ends and melt

Landing Gear

The fwd-rev retrofit solved the reverse castering problem, though it’s a somewhat painful transition. It would be most pleasant to have an EASY way to take the load off the front wheels... if done well, this could be a service stand, parking brake, and general stabilizer. In this general category, it’s fairly clear that we won’t get the deployable chocks done, which is a shame. We do need a way to keep the boat from careening out of control in land mode.
Lube main steering pulley bearings -- rusty!
Re-tension stretched deployment cables
Check bearings for saltwater incursion (hoist boat and exercise all systems)

Hydraulic System, General

This appears to be quite stable now.  After rudder kickup leak fix and 2 years of outgassing, do a purge-refill.
Repair leak in rudder kickup system - Teflon tape pressure relief valve. 
Full bleed and purge nasty reservoir; LG circuits are mushy
Improve reservoir fill system
Design autopilot interface: leadscrew/motor, 6x.75" cylinder, valve into system
Update the hydraulic schematic in DesignWorks (later)

Hatch Hinge Retrofit (later)

The forward and aft hatches seal well, but are a pain to use: they are free-floating and held down by straps and buckles. I really want them hinged, if it can be done without disrupting anything else.
Confirm hatch hinge design with spacers and minimal buildups
Bond/pot hinge mounting pads
Assemble hatch hinges
Pot internal deck padeye for locking line
Install blocks for locking line (2 on hatch so line not in the way)
Install fairlead in bulkhead for locking line
Install camcleats for locking lines
For gasket compression, 2:1 with 2 blocks on hatch to elim line interference
Remove straps and D-rings
Install utility hold-down bungees in old bowties if possible (only 1/8” clearance)

Solar Array, Power, and Thruster

Electric Thruster

This is going to be a key part of the system once the new solar panels are in place, and needs to be optimized.
Test new props underway, speed vs thruster current
    Install new prop in lab; test balance; package spare
    After on-water test, trim may be necessary
Hi-current cabling
Bypass switch for emergencies (currently only manual anyway; add temporary controller)

Basic Electrical Issues

General power distribution only; electronics are covered in the next section. Simple marine 12V and shore power are included here (and are outside the sealed enclosure anyway).
Add cockpit AC outlet to bring shore power to chargers and such (later, when console packaged)
Replace charger (it sucks 30mA when not plugged in), or add relay to enable only when on shore power
Confirm Group 27 battery OK; replace if needed
Add simple solar charge controller for small panel
Add main/backup change-over switch for survival loads
The distribution system might have to move (see console packaging)


This falls into two categories: navigation lights and interior/utility lighting. In general, we want to do it all with high-brightness LEDs to maximize efficiency.
Luxeon-based LED assemblies:
Acquire first four driver boards from George
    White emitter retrofit in sternlight
    Red and green LED retrofits in bow lights
    Deal with RFI issues from LED driver boards.  Different drivers?
    What happened to sternlight?
    All-round flashable anchor light
    Three white utility light assemblies (console, arch, and inside cowling) (later)
    Consider 5W Luxeon in Az-El steerable assy with IR, laser, and the two cameras (later)

New Solar Array

This is a big project: the existing ones, vacuum-bagged assemblies based on 1-inch foam cores, get too hot. We have to do this over, but can implement some improvements in the process.  This is critical prior to next mini-expedition.  It is instructive when trying to figure out the correct approach to consult the essential features necessary (besides the obvious maximization of power output):  center segments have to retract to allow access to landing gear deployment levers and clear Spinfin prop, a nonskid path to the dock has to exist, any part of the array should be able to take body weight (even though I'll try to avoid it), it has to be removable and stowable for land transport, there should be no finger/toe pinchers, it has to accommodate the twist of the ama relative to the center hull, it has to be partially retractible to minimize windage and wave impact, and of course it mustn't get too hot (as the current ones do, because of the foam core).  The following list assumes new aluminum panels, but I'm fretting about their potential to be really ugly.
Finish discussing structural design with David Berkstresser
Design outboard mounting
Acquire .875" OD stainless rail and hinge/flange fixtures
Install between aka ends, adjust aft riser to be parallel with gunwale (tweakable with trim)
    Forward mount vertical, leave 6" or so for ant/cam/light mount.  Safety cap.
Acquire one sample .090 sheet, as well as test channel
Channel stock: determine size, possibly 1x2x1, .080 or .125 wall
Select and acquire aluminum epoxy
Bond channels on edge (C) every 8 inches (4 per panel):  [     ]     [     ]
Fab test unit and measure deflection from body weight
Determine doubler geometry at each end, as well as any other fixturing
Walkway on backs of #3 units, each side
Punch and use grommets for cable routing
Soften all edges with nonwoven abrasive
Anodize before epoxy assembly; no dichromate (consider just alodining on-site)
Fabric hinges of Spandura; sample on hand
Connectors on pigtails that stow in channels. Add line for security sensor loop and temp probes
Stainless inserts for retainer pins; same folding protocol
Bungee hold-downs: loopback on 1 and 4; angled on 2 and 3 for retention in either state
NOTE:  Meeting scheduled for 7/17 with Rick Gannon... design may change

Songline Jobs

Songline is the sister-ship of Wordplay.  This is a preliminary list of tasks required to get this boat on the water.


Clean accumulated leaves, dirt, and standing water

Pedals and Seacycle Drive Unit

    R&R drive unit, clean rust, new oil bath
    Fabricate decent pedals
    Improve downhaul
    Replace chain
    Inspect to reduce static friction


    Fabricate aft dodger
    Re-do top todger, including stiffening rib
    Side windows
    Boom tent (for both boats)


    Fabricate double drum:
        Doorskin disc
        Foam-core disc
        Epoxy assembly
        Fiberglass and paint
        Add blocks and line guides as needed


    Anchor line (Ankarolina?)

Random Jobs

    Towing harness
    Pad steering T-handles
    De-wobble T-handles with proper Rulon bearings

Landing Gear

    New Stainless steering handle
    Bow wheel Retraction "gin pole"
    Change aft wheel line to Sta-set X
    Better aft retraction lines


    Identify all leaking or goo-damaged fittings and replace


    Daggerboard trunk leaks
    Improve hatch cover gasketing
    Fill puddling deck section under arch


    Fairleads for hatch cover lines
    Change furler and outhaul line to 1/4"
    Bow line improvement
    Bags for docklines and mainsheet


    Replace fiberglass tubing with stainless
    Modify seat positioning as needed


    Build subpanel on cowling for radio, switches, and basic instruments
    Rewire completely
    New battery

The remaining sections, still grayed-out, are all changing.  The system design has been completely refactored.

System Hardware and Packaging (later)

General comments: In the early years of this project, I assumed that there would be a complex, multifaceted, pressurized enclosure with lots of hinged gasketed panels -- a fabrication nightmare that would let me shoehorn techy bits into every cubic inch of space between the pedaling envelope and the cowling. Frankly, a project of this scale would be insane and doomed to leakage (recognition of which led me to first postulate active pressurization with desiccant-treated air, and then a bladder-wall that would prevent it ever pulling a vacuum). But people have built sealed boxes before, and if I could get away with less hardware, then I could just spend a couple hundred bucks and use something that has six sides instead of, um, 20 or so.

Fortunately, we can do more with less hardware these days, and probably even stuff it into a nice off-the-shelf NEMA-12 enclosure mounted forward of the pedaling envelope. The top of that space would then become a compartment for cruising guides and other essential reference materials, and the sides would allow stowage of long things. The display panel, a thin sealed box filled with LCDs and a hacked laptop, would hinge down to allow access to the compartment (while conveniently allowing use of the systems whilst supine). See? All that waiting paid off. That's what I was doing all year, just putting the project on hold so the industry's packing density could catch up with my daydreams. Really.

OK, so given all that, the question becomes what, exactly, is going to go in there? Some things are already known: the HF/SSB radio (which one?), satellite phone, power management hardware, some dedicated RF units, and various "appliances" that can't be replicated handily with software. The original crossbar assemblies are under scrutiny; audio and video are still pretty useful, but serial can be handled with a bunch of USB port expanders. I'll probably take an off-the-shelf marine chartplotter and mount it on the panel (simple, reliable, usable in bright light, and expandable at a stiff price with downloadable charts). A few dedicated LCDs will handle power system status (always on), video, and no doubt a few other things. But the question that keeps coming up in all the brainstorming here is what we need in terms of computing horsepower.

The first step is to list the requisite tasks. While underway, I need to do the following:

- Look at any sensor channel in real time or historically
- View an arbitrary group of channels on the same timebase
- Write with a simple text editor
- Read and respond to email, including attachments
- Schedule automatic telemetry transmissions
- APRS tracking (possibly standalone)
- Web browsing (limited by bandwidth and display constraints)
- Play MP3s from the library
- Capture images from the digital camera
- Record audio or video from any source
- Route A/V channels or invoke pre-defined setups
- Actively control connected subsystems as needed
- Operate the ham radio via remote commands
- Get range and bearing to other nodes in the flotilla
- Interact (voice/data) with other boats or backpacks
- Invoke security with some set of sensors and responses
- Hack the tools without needing an additional development system
- Have access to the full file system (PDFs, docs, photos, etc)
- Use various databases (people, projects, places, whatever)
- Trivial networking via 802.11 with manpack laptop
- Backup navigation (GPS, charting, CD->cartridge download)

A lot of this is off-the-shelf stuff, but a few issues arise immediately: text entry, cursor control, and display -- all on the general usability scale of a laptop or desktop system. There are various keyboard and mouse solutions that can handle the harsh environment if I'm willing to accept trade-offs in ease of use, but the display problem is a big one, as is power dissipation.

In the past year or so, some curves somewhere crossed and it became reasonable to think of a single machine, on the scale of an off-the-shelf laptop, that's always on... instead of our old model of occasional "big iron" and constant "hub." That was a nightmare of synchronization and focus, and I'm glad we didn't try to do it. But we still have a problem... a hacked laptop, like an iBook folded back on itself, mounted in the display enclosure, and magically cooled, becomes all but useless on sunny days. We're envisioning Squeak as a development environment for the "Microship App," and it likes a big bitmap... it would hardly do to require a viewing hood or squirt video out the side to a power-hog of a high-brightness LCD just to look at a virtual control panel. Even low-power wearable displays work better if your pupils aren't compressed to pinpoints by a tropical sun in your field of view.

It is conceivable that an external monochrome transflective LCD could be induced via USB to handle some of the display issues, but that puts us in the unpleasant position of having to write another GUI... and wheel-reinvention is anathema to an understaffed, underfunded project. Perhaps instead we can take something like the new Toshiba e740, a palmtop with bright display and lots of resources, and hang it on the same network with the Mac. With some clever hacking, we can move the active display space around as conditions dictate, saving the jobs that need a megapixel display environment for evening unless I don't mind scrolling around a virtual desktop.

So those are the big design issues at the moment. We now return to our regularly scheduled TO-DO list.

Console Packaging Project

Though it may break out into new subcategories later, this consists of the pedaling envelope, the display enclosure, and the system box.
Move power distribution stuff to port side of pedaling envelope
Fill old holes
Make template of console display space to begin choosing LCDs
Is there something suitable off-the-shelf for the display enclosure?
Glass envelope to deck and remove the aluminum angles
Select largest available NEMA enclosure (that will fit) with clear top, possibly 2 linked by tubing.  9x12x32 approximately
Decide re standalone chartplotter vs software for palmtop (affects enclosure)
Get rid of olive-drab switchbox (poor function-to-weight ratio)
Build switch panel (port side) after determining needs, with splash/foot cover
Install vapor-phase corrosion inhibitors in all enclosures
Provide a Gore-Tex breather plug, or more likely a bladder, to prevent pulling a vacuum

Computer and Control Systems

This, as noted above, is still a bit on the vague side. The to-do list reflects that. It’s fairly critical that we make these decisions soon. In general, we want to minimize the stand-alone packages from the earlier design; speech, for example, really should be from system software, not a dedicated board. Be very discriminating here... keep the hardware as simple as possible.
Set up "simulator" area in lab:
    Clean system corner that has become a boneyard
    Clear static documentation and other relics from adjacent shelving
    Install new 8' fluorescent
    Install Linux box, with current Squeak version
    Clean up crossbars
    Dedicate Bionode to A/V crossbar control
    Acquire some more USB-serial interfaces (DLP), hubs, shorty cables, etc.
    Breadboard additional Bionodes for general I/O
    Set up all known on-board devices
    Order/acquire all sensors
    Identify channels and write reference document
    Cabling with an eye toward packaging in system enclosure
    Start writing front-end tools
Catalog all I/O from the box to home in on waterproof connector requirements
Build up 4 more Bionode boards
Acquire about a dozen DLP USB-RS232 boards
Find solution to high-ambient display problem that doesn't suck too much power
Build new Auxbar/Vixbar controller with Bionode front end
Build bank of power controllers (SSRs and FETs), with Bionode front end
Change Auxbar op-amps to low-power units (find and order pin-compatible parts to replace TL074)
Replicate dated diagnostic LED matrix display with discretes (or other active "virtual front panel")

Sensor Channels

These aren’t quite laid out as to-do items, but all have to be included and brought back to the various real-world interfaces hanging off the Hub’s USB port. Some of these are more critical than others. Some are essential for security or navigation; others are a bit on the blue sky side.
Acquire all these:

Radiation (pulse count)
RF field strength (xmtrs, broadband or sweep)
EM field strength (60 Hz)
Temp/humidity/light level (analog 0-5)
Barometric pressure (analog)
Wind direction/speed (serial I/O)
Compass (serial Ritchie gimbal unit or Etak analog flux-gate)
GPS (serial)
Internal temperatures: batt, console, hub system, cockpit (analog)
Solar array temperatures if cabling permits
Power subsystem telemetry (serial)
Multichannel water probe (serial)
Water temp, salinity, pH, turbidity, conductivity, depth
Security bits (magnetic sensors, switches, interlocks, LG status, etc)
Bilgewater depth (or at least presence); all 5 regions
Shock/vibration/acceleration/tilt sensors
Proximity sensors for security (ultrasonic)
Microwave motion sensor
IR motion sensor (internal to cockpit)

Communication Systems

Like power, this tends to overlap with a lot of other things. This section is mostly to identify the major comm devices for purposes of hardware selection and packaging.
APRS standalone with PIC-E and dedicated rig (790?) mostly for flotilla use
Incorporate back-door control functionality via packet through 790
APRS full-featured with HF. An Elecraft K2 may be appropriate here.
General HF functionality. Icom super-rig, or K2 for low power?
If K2, get one and get busy... it’s a kit
Icom 706 can be home-base telemetry system
KAM+ on serial/audio network
G* satellite integration (talk with Qcom re currency and account)
Antenna tuner, latching relays essential
HF antenna selection
Grounding/counterpoise system: salt/fresh/land
Marine VHF integration, but it is standalone for safety
GMRS/FRS built in to simplify comms with flotilla handhelds
Standard cellular phone (car kit, ext antenna); choose service. Ant is 900/2.4
AM/FM receiver
FM synthesized transmitter for group programming (yeah, DMCA, I know)
Video transmitter and receiver
Cable GPS to Marine VHF for DSC functionality
FM xmtr on xbar with digital channelized waterproof headphone walkman

Software Projects

I have just finished listing the Squeak applications that will run in the connected client systems (console PDA, remote laptop, etc), as well as a complete I/O catalog for the Linux hub itself.  We're now looking at the hardware choice for this system, and will use Spread as a sort of software crossbar network, talking via USB with distributed homebrew PIC I/O devices called Bionodes. This should begin to translate into specific software projects soon; Ned Konz is currently working on the infrastructure tools.  The skeletal documentation for the Shipnet, quite out of date, is here.