By now we were well-established at UCSD, with student teams cranking away on projects and serious work happening in the marine architecture department. We’re about to switch to each Microship Status Report having its own post in these archives, but this is all of the ones posted to interested parties in January ’94.
Microship Status 1/7/94
by Steven K. Roberts
In This Issue:
- ECE Project Teams
- MCS Hub Manager
- Audio Crossbar Network
- Serial Crossbar Network
- Video Switcher
- Environmental Data Collection
- Gps And Initial Navigation Sensors
- Ultrasonic Ranging System
- Solar Array Profiler
- Hull And Deck Weight Estimates
Distribution of these “daily reports” is currently 110 people: 68 UCSD students, 8 UCSD faculty/staff, 34 industry, sponsors, friends, and others
Yikes. It’s getting harder and harder to keep up with these reports — I’m finally driven to writing one when the accumulated activities make me realize that it’s either going to be embarrassingly sketchy or ponderously huge, and that I better take a break long enough to report what I’ve been up to. Things have been BUSY here.
First, a quick scheduling update — I mentioned in the 12/27/93 issue that I might be going to Seattle/Victoria this month to pick up the new kayak outriggers. I’ve decided against this, since the ECE190B class is now in full swing and being gone for 2 of the remaining 9 weeks is probably a bad idea. So I’ll be around almost every day, probably as a succession of marathons like the past 60 hours… working til I drop, sleeping in the lab, crawling to the basement shower, then diving back in… zombie-like, driven, caffeine-wired, obsessive, immersed in FORTH/proposals/meetings/specs/hardware. But hey, tonight I go HOME for the evening! I find that strangely exciting…
Anyway… on with the updates…
ECE Project Teams
At present count, there are 12 people in the ECE190B senior design class with Clark Guest and me, working on the ship. Also, all learning FORTH, 68HC11 interfacing, multidrop networking, packaging, documentation, and, and…
The following paragraphs briefly detail the projects that have been assigned to or chosen by the participants. These do not necessarily correspond one-to-one with nodes on the control network — the objective is to identify conceptual components that are on a realistic scale for a successful one-quarter project, and package them in a way that can be easily integrated with later work. Nor does this represent
ALL of the projects that have been identified — limits on available people, space, tools, and hardware force us to put off a few things till Spring.
MCS HUB MANAGER (Michael Bream and Chris Tuft). This is the central FORTH software for the Hub, the 68HC11 that sits atop the control network of 15 other processors. This project includes the basic monitoring functions, Easy-A protocol tools, 2-line LCD status display, simple keypad access, pass-through connections from a high-level application system, and inter-node mail processing that allows a sort of poor man’s peer-to-peer networking.\
AUDIO CROSSBAR NETWORK (Isaac Chu and Jason Corley). This system owns a massive 8816-based analog crossbar network with 32 inputs and 32 outputs, all conditioned by TL084 op amps to allow standard line-levels through the net. All of the ship’s audio devices will interconnect through this, and the project will yield the device control tools, internal array that reflects crossbar status, and a text-based front-end that allows high-level systems (or humans) to add and drop links. The team will also subject the finished network to audio noise and distortion tests, and provide initial calibration for the test suite of devices in the lab.
SERIAL CROSSBAR NETWORK (Jeff Simon and Dan Sebald). Architecturally, this is similar to the above, but all switching is done in pairs to accommodate the needs of random serial interconnects throughout the ship. The software tools locally provide all the network management, and also include some functions in the Hub that allow it to intelligently re-route its own console and permit cellular logins or alternate host systems. One special function allows any high-level processor to “subscribe” to the multidrop network traffic for diagnostic purposes.
VIDEO SWITCHER (Delon Levi). This project will experimentally establish whether 8816 bandwidth is sufficient to pass video signals without distortion, and will proceed to implement a crossbar network of 16×8 nodes. Later, this will accommodate the ship’s cameras, still-frame processing, ATV transceiver, console monitor, and so on. During this quarter, the objective here will be to selectively route a couple of available video sources to the Ampro/Sharp color LCD display or to a local VCR/monitor. Luke Abbott, who is our official project videographer, will work with Delon Levi on this.
ENVIRONMENTAL DATA COLLECTION (Tom Barnes). Although we do not yet have the full suite of environmental probes (which will hopefully include temperature, pH, salinity, turbidity, dissolved oxygen, radiation, and various meteorological sensors), this project will begin with radiation and temperature and hopefully add one or two water-quality sensors from Scripps. The software will acquire data blocks on demand, tag them with a time stamp as well as latitude and longitude from the GPS receiver, and return a variable-length string to the host.
GPS AND INITIAL NAVIGATION SENSORS (Henry Xia). This system, which will expand into a fully autonomous navigation role, will begin this quarter with tools to acquire GPS data from the Motorola Traxar + (NMEA 0183) as well as depth and compass heading. These are maintained in an internal array, and are retrieved every second or so by the hub for display on the console LCD. Next quarter, we will try to integrate this with packet radio and GeoQuery on the Macintosh, allowing a person to walk around the UCSD campus and be tracked by a moving blip on an electronic map.
ULTRASONIC RANGING SYSTEM (Carl Mascarenas). This began as a component in the ship’s security system, but may have application as well in foggy docking and “watch mode” when knowing the distance to an object well inside minimum radar range is valuable. This project will reverse-engineer a commercial ultrasonic ranging device, extract BCD distance from LED drive signals, and return the distance to the nearest object whenever requested by the Hub via the control network.
SOLAR ARRAY PROFILER (Mohammed Sarhadi and Mehrdad Heydari). In order to allow ongoing diagnostics and module performance analysis of the solar panels, this project will yield a pair of processors, one for each array, that maintain internal running averages of the current sourced by each of the 12 modules in its local domain. Over time, if trouble occurs, these values will reveal declining performance that would otherwise be impossible to assess in a parallel-connected system. Via inter-node mail, these processors also enable or disable their 360-watt charge sources onto the bus and report the amount of current they are delivering.
As all these develop over the next 9 weeks, we’ll issue summary reports on progress here (though much more detail will be available in the mailing list limited to the participants).
Other projects are underway in Computer Science and Mechanical Engineering, and I’ll report on those next time. These eight will be keeping me busy for a while, as we are also building the bench network, learning FORTH together, and dealing with a wide variety of interface issues. It’s going to be an exhausing 9 weeks, but when it’s all over, the Microship Control System should be well established, easily extensible, and hackable from a remote host. Among many other things, we’re looking at four different applications for fuzzy logic, Ampro front-end software, Mac wireless networking, cellular phone interfaces, ham and marine radio control, the video turret with gyroscopic platform reference, battery management, thruster control, and much, much more… Stay Tuned!
You might remember from my previous mutterings in these reports that I spent most of the “holidays” writing the initial specs on the control system. This 55-page document is now complete (it will change, of course), and you’re welcome to peruse it at your liesure. I will probably maintain the current version on the NRL.UCSD.EDU gopher site, which John Studarus has implemented on the Tadpole SPARCbook. At the moment, though, you have your choice of hardcopy in my office or email in yours…
Hull And Deck Weight Estimates
Robb Walker has just reported the following:
Preliminary estimates for main hull/deck shell (no structure):
Hull Surface Area: 275 sq. ft. Wetted Surface: 95 sq. ft. Deck Surface Area: 100 sq. ft.
Base hull panel wt: 1.2 lbs/sq. ft. Bottom panel wt: 1.5 lbs/sq. ft.
Hull/Deck Shell Wt: 360 lbs. VCG: 2.60 ft. below deck (-2.60) LCG: 15 ft aft bow (approx. 5.0)
Bulkhead and primary structure weights can be estimated at 1.0 lbs/sq. ft. Secondary structure can be estimated at .5-.75 lbs/sq. ft.
Robb, as well as the mechanical group including TJ Tyler and Jeff Klompus (who will be working together this quarter on the structural analysis of crossbeams, trailering, and internal frame), will be meeting here at the lab tomorrow to discuss the structure in detail. I’ll report on this soon!
I’m fried. There’s more to tell, but I haven’t been more than a coffeshop away from the engineering building for three days. I’m outta here….
Microship Status 1/18/94
Lots Of Random Updates
Still alive! This is embarrassing — my last daily status report was 11 days ago. Things have been incredibly busy around here, with so much happening that the easiest way for me to catch up is with a bunch of random updates and no attempt at cohesion…
Today, I installed the Real Time Clock and the Serial boards on the Hub development system, and we can watch the time changing via the LCD with a few simple lines of FORTH code. This involved recycling a 34-pin ribbon connector (a rather ugly process), and mounting the boards on standoffs on the chassis. But it worked the first time, something that always warms a hardware hacker’s heart…
Dan Yang is now working here almost every day. He’s from Taiwan (a Taipei personality…), and is into systems & controls. He’ll be our “lab coordinator,” a job that includes asset relocation, fabrication projects, parts ordering, and learning everything about everyone’s project in order to provide continuity across quarter boundaries. Today, Dan rescued my vise from Seaweed Canyon and mounted it on the bench, reversed the RJ-11 connectors on one end of every inter-node cable, then installed RESET pushbuttons in all the control network junction boxes. All this adds up, given a bit more work, to a bench network for FORTH development that allows easy installation and disconnection of individual nodes.
Agnes deGuzman is also working here almost every day — she’s becoming the engineering project manager for 199 credit. She has just begun the learning curve, and is reading my articles and design specs, sending media copies to sponsors, and sorting out the vast library of help offers from dozens of people who would like to be involved. As some of you (hi, Jeremy) have doubtless noticed, I tend to be inefficient at following up on offers of project assistance, even though I’m often overloaded here. Much of this is simple difficulty in delegating until I know peoples’ skills; the rest is the half-megabyte of saved email that’s just too much to wade through. Agnes is dealing with it, and will help with all management aspects of the whole process…
Michael Bream and Chris Tuft upgraded the SE/30 to System 7 today, and helped clear out a bunch of old disk clutter. Feels wonderful to have consistent, high-quality tools across all platforms! Since this is the development console for all the control system projects (ECE-190B), it’s great to have the space clean and well-organized.
The FORTH projects are all developing in parallel, which is interesting to watch. I’ll post details here as appropriate, but in general, we’ve seen lots of FORTH education, agreement on central data structure between audio and serial crossbar teams, initial GPS familiarization and external power connection, and considerable research on all fronts. The lab has become a busy place…
John Studarus and the rest of the Information Networking Team have brought up a World Wide Web server on our Tadpole SPARCbook. Access it via Mosaic using the URL: //nrl.ucsd.edu/ and you’ll find images and full textual archives… all still under development. The public interest from Out There is amazing: yesterday, a holiday, saw 296 accesses from the Net.
Last week is a sort of blur — I was sick for the first half and overloaded the second. The Saturday previous included a worthwhile event, however — about 10 mechanical engineering students showed up here for a sort of orientation meeting led by TJ Tyler and Robb Walker. We identified four major project areas (overall frame structure, crossbeams, wheels, and outrigger attachments) and did quite a bit of useful brainstorming. TJ and Jeff Klompus have now been officially sanctioned by the Mechanical Engineering Department to do a technical elective on the frame structure, including Finite Element Analysis on the Cray C-90. We’ll have lots to report on this as it develops.
Clark Guest has written a FORTH reference stack in HyperCard, using the New Micros literature as a source. If you’re working on any of the control systems, this, a copy of Pocket FORTH for the Mac, and the second edition of the Brodie book are now the essential references for learning curve development.
Lots of care packages arrived last week! In addition to a batch of connectors and cables and the 13 FORTH boards from New Micros, we received the complete Ampro 286 system — an amazing assemblage of 3.6 x 3.8 inch stacking circuit boards. The boards include the CPU, Ethernet, VGA, and disk and I/O controller. There is also the interface card for the Sharp color LCD. We haven’t even started to put all this together yet — I’m looking for someone very DOS-literate who wants to take on both hardware and software for what will become the control system’s host environment.
Literature is flowing in, but I’ve given up on the “Literature Received” column that was a regular feature of these reports. Nobody ever seemed to use it, not even me. I will say that we’ve been receiving a fairly steady diet of magazines, catalogs, and spec sheets in response to a number of simultaneous research projects into such things as autopilots, clutches, composites, boating accessories, and more. If you need information on any part of this, check the library here… Agnes and I have its organization near the top of the TO-DO list.
Much is afoot, but I’ll leave it at this quick update for now — more detailed system close-ups will follow.
Microship Status 1/19/94
In This Issue:
- Digital Ocean Wireless Appletalk
- Mcs Network Hardware Status
- Gps Interface
- A Forth Real-time Clock Program
No major articles today, just more quick status updates. I realized I was getting into a feature-writing mode, treating this as a forum for relatively long and coherent pieces which are hard to get around to doing. That’s what the Nomadness Notes and the corresponding hardcopy newsletter are for. This series of project status reports, on the other hand, should take the form of timely and relevant news of interest to anyone close to the project, with only occasional tutorials and full-scale articles.
A discussion with Digital Ocean (Overland Park, KS) revealed the happy news that they are close to shipping their Grouper product, a very cool wireless AppleTalk link. We’ll have a pair of these on the ship to allow the PowerBook to be present on the ShipNet while off-boat. Initial experimentation (around April) will use their current product, which assumes LocalTalk connection to the Mac host; later we’ll replace that with a new unit that will live on the Ethernet (the ship’s networking backbone).
IEE (waterproof keyboards) is down due to the quake, and I left a message for Sharp (color LCD for the Ampro system). Frank Araullo is planning to bring up the Ampro, and we need to borrow a video monitor from someone until the LCD arrives.
Dan Yang ventured into the world today, picking up 100′ of RG-8X coaxial cable from Ham Radio Outlet and the SIP connectors for the FORTH boards from Gateway. Agnes ordered the power connectors from Mouser, and I’m spending time this evening hardware-configuring the NMIS-5000 serial board that will let the Hub processor talk to the multidrop network via modified RS-422.
Henry Xia and Dan worked with the Motorola Traxar GPS for quite a while today, eventually getting it interfaced (4800 baud) with the Zenith laptop and quite happily displaying navigation sentences at a 1-second rate. Of course, the data in the lab gets a bit boring after a while, but at least Henry can now start working on sentence parsing and so on. We’ll need to interface a serial board to his FORTH node.
Delon Levi and Luke Abbott met here tonight to discuss the video crossbar requirements, and are now trying to figure out what the standard video signal level/format through the net should be (video only, NTSC, or whatever). I welcome suggestions from anyone who has dealt with video at this level. Delon contacted Mitel today (makers of 8816s and other crosspoint chips) and discovered that they have a new model (88V32) that has internal ground paths and other crosstalk-minimizing features that enable it to work well with video. Samples are enroute.
A Forth Real-time Clock Program
I spent a few more minutes playing with the real-time clock on the Hub, and wrote a simple program that displays the time on the 2-line LCD:
: DIGIT C@ 30 + DSP-EMIT ;
: SECS C001 DIGIT C000 DIGIT ;
: MINS C003 DIGIT C002 DIGIT ;
: HOURS C005 DIGIT C004 DIGIT ;
: Colon 4a Dsp-emit ;
: Time Hours Colon Mins Colon Secs ;
: DELAY 1000 0 DO LOOP ;
: CLOCK BEGIN CLEAR TIME DELAY 0 UNTIL ;
DIGIT simply fetches the BCD values from the memory-mapped time buffers (starting at C000) and adds them to the base value for ASCII numeric digits.
Note that the “BEGIN xxxxx 0 UNTIL” is an easy way to create an endless loop, since the 0 on the stack always ensures that UNTIL will never be true. Hitting the processor’s RESET button is the only way to end the program short of the power switch, a hammer, large-scale static discharge, or physical chip removal.
What’s wrong with this as a programming example is that the memory addresses for the time values are embedded in inline code. It was just a quick interactive keyboard hack — a more elegant solution would be to define a base address with offsets, read the values with a loop, integrate the date, add setting and alarm features, etc. I leave these and other refinements (such as display in Katakana text strings) as exercises for the reader…
Ain’t technology wonderful?
Microship Status 1/20/94
In This Issue:
- Progress On MCS Network
- Other Updates
A good, busy day in the lab. Today was the meeting of the ECE class of 11 people working on FORTH controller projects, and took the form of a very spirited discussion about data structures, protocols, and the specifics of everyone’s machines. It’s getting interesting…
Progress On MCS Network
Most of today’s hands-on effort involved the multidrop network, with Michael Bream, Chris Tuft, Dan Yang, and me working together. Michael distinguished himself with some very clean packaging work (he has spent years installing car stereo, cellular, and security products as well as doing various other fabrication projects, and it shows). I recommend that anyone who’s working on Microship electronics projects without prior hardware experience spend some time with Michael to pick up some pointers on cabling, packaging, soldering, and otherwise doing hands-on electronics… it will pay off.
Anyway, the task at hand was to configure and bring up the serial board on the Hub, make the Hub’s network interface box, and cable at least one node to the net. All of this takes an amazing amount of time, requiring the hacking of not-quite-right connectors, finding cable, and debugging the whole mess — a task complicated by the garbage RJ-11 junction boxes we bought by mail order from All Electronics. About half of the insulation displacement connections turned out to be intermittent or open, easily explaining why they were so cheap on the surplus market. Now we have to pay with time instead of money, soldering all the connections by hand.
Results? We’re not there yet. A FORTH node is now connected physically to the Hub, and we can see data flying around when we write a transmit loop, but we have not yet been able to see a receive data flag indicating flow in the other direction. Once we get that going and write some basic tools, we’ll start programming the on-chip EEPROM in each board with its assigned net address, and, hopefully, find a way to install some non-volatile Easy-A code so a crashed node won’t take down the whole network. Ah, learning curves…
By the way, a little aside for anyone working with the Hub development system: the Power One 5-volt supply on the chassis can go into crowbar mode when briefly shorted, and will STAY that way until the unit is physically unplugged from the AC line and allowed to cool.
Another thing about power: I’m installing screw terminals on the FORTH boards, which can be wired backwards if you’re careless. PLEASE don’t do this, but if you do, I’m also installing a 1N4001 protection diode on each board, reverse-connected across power. This will smoke profusely while causing severe embarrassment and loss of style points. Please be VERY careful with all wiring if you are doing any kind of hands-on work here! (The looped diode anode lead, by the way, provides a handy ground point for a scope probe.)
I’ll ask TJ Tyler and Jeff Klompus to write a whole report about this when they’re at a good checkpoint in the structure project, but today they spent about 3 hours with my hot glue gun and a bunch of toothpicks left over from mini hot-air balloon projects. There is now a miniature model of the Microship internal frame (rev 0.1) that allows visceral analysis of stress modes. While we all love CAD and other design tools, it really helps to make models that can be handled and passed around — it’s a tremendous aid to visualization and understanding.
Agnes shipped copies of recent media coverage to all named sponsors today, bringing us up to date with this never-ending but pleasant task. The latest addition to the media collection arrived just in time today — the Winter 93-94 issue of UCSD Perspectives, mailed to alumni and contributors, containing a good 3-page article about the Microship and BEHEMOTH.
Agnes is also working on the rather daunting task of figuring out how to match the growing project list to available human resources. Those of you who have submitted questionnaire responses are already in good shape — if you have not done so and need a copy of the 10 questions I circulated to all student volunteers before Christmas, please email her directly.
I spoke with Kirk Elder and Bill Lee, two wizards in high-efficiency electric motor/generators. The next step in thruster design is to track down information on units from “Unique Mobility” in Colorado — reputedly up to 98% efficient with sumerium cobalt magnets, and the motors of choice for most of the solar race cars. Kirk likes (and sells) GE series-wound traction motors; Bill sells variable reluctance and prefers those. Much of engineering lies in sorting out the varying opinions about all choices…
We’ll have a very welcome guest in the lab this weekend — Bill Brown, WB8ELK. He’s the former editor of 73 Magazine, and is a consummate hardware hacker with years of experience in ham radio, video, instrumented high-altitude balloon, voice IDers, and control systems. I’m sure some new ideas will be circulating within the next few days.
Enough for now — I’m going to go pound the FORTH keyboard and see if I can get the Hub talking to one of the remotes! I need a win before bed…
Microship Status 1/25/94
In This Issue:
- Hub-node Communication Progress
- Kayak Completion
- Random Notes
Hub-node Communication Progress
Well, we’re into the hacking for real. Proving once again that nothing is trivial where electronics and computers are involved, we’ve been spending days reverse-engineering the network connections and related control circuitry on the FORTH boards, talking to various experts, trying hardware hacks, writing test loops, staring at the scope, scratching our heads, correcting documentation errors, and otherwise going through the steps necessary to awaken the network.
In the last issue, I reported that we had not yet established Hub-Node communication. We’re well past that now, with a good tool for accessing any single remote processor connected to the Hub’s serial port. Bill Muench, a FORTH guru in Santa Cruz, sent a new tool for high-speed uploads via a handshaking protocol with the Mac, and I wrote a corresponding MicroPhone script. The learning curve is proceeding well, and I’m sure we’ll get there.
But it turns out that the Easy-A protocol, as written, leaves the currently unselected boards in what amounts to a WAIT state, meaning that they cannot be out there accomplishing useful work unless they are talking to the Hub. So it appears that we’re going to need either a simple multitasker, or some clever interrupt routines linked to the serial port that check all passing net traffic for the switch character that means a given board is suddenly in the limelight.
Michael Bream and David Wright (an old friend/wiz visiting enroute to Baja) have been working together on the basic problem of getting Easy-A to work for much of the evening, and have made considerable progress involving some hardware hacks. I’ll report more fully when this is actually done, but there’s some very strange stuff going on in there and we’re not out of the woods yet. If you’re working on a project and are on hold pending access to a board, please bear with me for a few more days, do as much development as possible with Pocket FORTH, and work on your hardware design.
In related news, the network node boxes for the development environment are done, and the power connectors and protection diodes are all on. Michael cleaned up the 5V power supply, and I made a binder of hardcopy documentation including the FORTH memory map, a great 68HC11 I/O spec, and processor board details. Students in the ECE-190B class are making considerable progress on various projects (more on that soon), and the lab is no longer a quiet place until late at night… when I can at last turn down the lights, play some music, and catch up on email.
Finally, speaking of computers, Frank Araullo and Dan Yang have begun the packaging of the Ampro system, and hopefully will be bringing it online in the next few days. Still no LCD or Hub keypad.
The outrigger kayaks have been completed by Current Designs in British Columbia, and will be driven down by Bobbi Smith in a couple of weeks. She sent a summary of the technical description of these custom craft:
Normally the standard laminate is: – one 18 ounce roving – one cut down 15 ounce roving – one full 18 ounce roving
What they did differently was to use: – one 18 ounce roving – one full 15 ounce Kevlar hybrid – one 18 ounce woven roving (stiffer laminate in the hulls of the boats replaced a layer of 18 ounce roving)
- plus added some reinforcements (that aren’t standard) down the keel line (15 ounce woven roving reinforcements)
- also in the laminate is a upicamat core running longitudinally down the boat’s center line through the flatter areas of the hull to stiffen the hull (always standard in either fibreglass or the kevlar)
Also included as far as parts in the shipment: – extrusions for joining the hull and deck (currently screwed together with flanges) – fibreglass seam material to join hull and deck – front and rear bulk-head – front and rear hatch systems – four tubes of kop-r-lastic for the extrusion when joining the boats as well as for bonding in the bulk-heads
I’ve been doing a bit more research on motors for the thrusters, and found that the GE series brush units from the Florida outfit are 1 horsepower, 25 pounds, 8″ diameter, and 12″ long. This is HUGE and quite absurd. I’ll look elsewhere…
An order is going in to Digi-Key tomorrow for LM10 op-amps and LM335 temperature sensors. If you have other items on the wish list, please let me know ASAP!
Bill Brown was here this weekend, and we’re discussing a partnership on the BEHEMOTH Technical Manual project, which will never get done by me alone.
Mitel sponsored three 88V32 video crosspoint chips (Delon Levi’s project). The bad news is that they are in PLCC packages, so we’ll need Emulation Technology adapters for testing .
I received data on over-running clutches that may eliminate the need for power-hungry electrical clutches between pedals and thrusters.
Finally, I did an interview today with the San Diego Union… should be out in a few days. There will be a photo session in the lab Thursday afternoon at 3:00, so if you’re in the ECE class you might have a moment of fame ‘n glory…
Time to shut down for the night. Cheers!
High-level Systems and Applications
Macintosh. The primary working environment will be a Mac PowerBook, probably the 180C with color active-matrix screen. This needs to be packaged, fed with synthesized ADB console devices, linked to serial crosspoint and network interfaces, and provided with power management. -> Physical repackaging in sealed console enclosure -> Mouse interface from DOS-intended or custom FSR pointer -> Keyboard ADB hack (again, hardware designed for DOS) -> Network hooks to other console Mac and RF to laptop -> Replacement of normal battery/charger with system power -> Installation of general software tools -> Hooks to Internet, Satcom driver, other email tools -> Backup tape or file server in equipment bay; CDROM drive
Navigation system manager, integrating GPS receiver, flux-gate compass, waypoints, chart display, etc. The NMEA 0183 bus will be connected to one of the high-level systems (probably the 486, keeping the Mac free for personal productivity, writing, and email). The nav system will integrate live charting with the contact database while providing the full suite of tools needed at the helm of a well-equipped cruising yacht (including support for celestial nav). -> Repackaging of 486 (probably Compaq) -> NMEA interface -> Select and acquire nav/charting/marine software -> Network to Mac -> OrCAD and/or other DOS CAD tools -> Power management -> Integration of KB and Mouse with Mac group
Inter-boat and boat-manpack RF networking. The Mac on the console will be in constant data communication (AppleTalk) with companion craft, the copilot console, and the manpack laptop. A recent visit to Digital Ocean has suggested that their spread-spectrum “Grouper” product is ideal for this. We need to consider antenna issues as well as integration with a hardwired net. This is an area with extensive industry spinoff potential — wireless data is a very hot field. -> Acquire Digital Ocean equipment and test -> Integrate with wired net to other Mac -> Establish simple protocol for other boats -> Gateway software to make Microship email hub
Human interface issues. Chord or waterproof QWERTY keyboard, macro expansion tool for the Mac, cursor control (probably InterLink force-sensing resistor based), speech, and hardware/software solutions for seamlessly sharing these resources among the three high-level systems. -> Compare InterLink DuraPoint with FSR XYZ custom unit -> Write ADB driver for above (whichever wins) -> Make (Paravant?) sealed kbd interface switchable Mac-DOS-DOS -> Explore speech input options -> Integrate Audapter (or?) with MCS hub and xbars -> Software front-end for MCS display manager
Video and conferencing system. Frame-grabber, compression, control of on-board cameras and other sources, transmission of stills via cellular and satellite email, motion using Colby and other compression schemes, multimedia applications on-board, CuSeeME conferencing via the Net, and more. -> Track down Cohu or similar small cameras -> Investigate available frame-grabbers -> Choose video platform -> Talk with Colby re support for compression -> Try CuSeeME via cellular -> Define and integrate video sources/sinks via crosspoint switch -> Software for inclusion of JPEG images with hourly postings
Geographic Information System integrated with database and GPS, built around GeoQuery. Currently we need Mac drivers for the GPS to automate map centering around the ship’s current location, and the software is not yet tested with commercially available machine-readable charts (only the internal road and state/county boundary maps). All database entries are geocoded and presented in graphic context. -> Get GeoQuery and ComGrafix together -> GPS data into Mac -> Select Mac/Dos platform for CD-based charts (or cartridge?) -> Investigate putting CD chart lib at MC sys and downloading -> Integration of chart data with radar/nav (commercial software) -> Build auxiliary databases from EYP and elsewhere; update code
Multimedia demonstration package to integrate everything into a demonstration, considering how often this will be on stage or in the media. Speech, graphics, active control of systems, and so on…
-> Managers: assure up front that hooks exist for this
On-board CAD tools. System maintenance underway should be facilitated by a hierachical file structure, accessible graphically, for all mechanical structures, electronic systems, pneumatics, water processing, sail controls, and so on. This should be designed up front to fall directly out of the project development tools.
Networking and Support Systems
The Mission Control system (map display, telemetry decode & distribution, mail hub, alarms). This might run on a SPARC, with fleshed out (human-readable and abstracted) data fingerable and appended to an ftp server in a variety of formats that support tabular download, geographic plotting, and trend analysis. It should be possible for anyone on the Net to finger this machine and receive a complete snapshot of Microship activity, including lat-long, heading, speed, depth, weather, time, air/water temps, radiation, salinity, turbidity, pH, solar performance, battery level, propulsion level, water system levels, security status, recent statistics, brief commentary, and so on. -> Establish hourly report format (work with NSF/TERC) -> Code to archive in useful form (chartable) -> Franklin Antonio re mapping (QTRACS) -> Converters as necessry for image formats -> Login shell for remote system maintenance and diagnostics
Satellite Internet link. I’m working closely with Qualcomm, an active sponsor, on this part. Issues yet to be decided include antenna packaging (microwave attenuation through fiberglass is apparently not too severe, and mounting below a deck bubble will minimize seawater damage), polarization tricks at the limits of the footprint, and smooth integration with on-board and Mission Control systems. We have already developed the code for BEHEMOTH to support Eudora on the front end and a SPARC task on the back end, effectively yielding a seamless email path anywhere in North America, Europe, the Pacific Rim, or Australia. This team should also study Standard-C satellite service and stay alert to the implications of the upcoming Iridium system. -> Select between two models of OmniTRACS -> Full test of MCTtool in the Mac -> Physical packaging of terminal and antenna units -> Study possible need for gimballing (hopefully avoided) — test! -> Write SPARC code to “throw switch” between sat and phone path -> Email filter -> Connect with wireless networking community re future
Cellular phone, Axcell, modems, and so on — This might fold into comm systems and networking, but the industry is heating up in its quest for people literate in wireless networking. Cellular phone integration and interfacing. -> Choose phone/modem combo (Motorola UDS or Axcell-supported) -> Talk to Qualcomm re dual-mode CDMA/analog unit -> Study new cellular data options… changing fast -> Physical repackaging of phone, interface hooks to net, antenna
Fabrication, Mechanical Design, and Packaging
The video turret — a waterproof glass dome on the deck, housing a video camera that can be steered and zoomed under software control. Perhaps night-vision can be integrated with this… -> Study existing commercial systems (oceanographic) -> Find outer envelope, sealed -> Choose camera, interfaceable -> Code for camera control -> Apps software (security, documentation, vessel ID)
Equipment packaging and environmental protection. Pressurized equipment bays (with dryer), connector choices, sealed boards, sensors for salt/water incursion, pressure drop, excessive temp rise, shock, and more. This group will be responsible for the air-handling system, valves, sealing protocols, connector choices, and more — coordinating other packaging teams to maintain consistency in this extremely critical area. Consulting is available from Scripps Institiution of Oceanography, MBARI, and some of the industry sponsors. -> Select pump, filter, desiccant, valves, and tubing/fitting standard -> Test system for cleanability if loaded with salt water -> Pressure sensors for each enclosure, ref to ambient -> Control software for pump and alarms (see controls, above) -> Temp, shock, door ajar, and other package monitors -> Set standards for connectors, bulkead fittings, sealing protocols
Hull design and stress analysis. Work with Nelson/Marek yacht design and our epoxy/composites sponsors to optimize hull shape, accommodate the cross-beam and landing gear stresses, and implement using the appropriate fabrication technique (probably either cylinder molding with compounded ply or the more traditional strip composite construction that accommodates more complex hull shapes). Finite-element analysis of complete structure. -> Express complete design in AutoCAD -> Pass CAD files and material specs to FEA group -> Choose fabrication method -> Set up vacuum-bagging facility -> Strong-back jig and station templates from N/M design data -> Stress analysis of bulkhead-area forces (wheels and akas) -> Design akas as sealed beams for flotation and strength -> Determine whether commercial double kayaks adequate -> Design bulkheads and internal structure -> Build hull(s) -> Cable routing and deck fixtures
Road mode. The Microship will contain integral trailering capability, with fold-down wheels from the bulkhead stations and a bolt-on hitch for the bowsprit. The stresses will be enormous, and up-front design must consider the shock amplitudes and frequencies encountered on the highway, point-loading stresses, suspension design, and more. -> Study laws and standards covering trailer design -> Implications of four widely-spaced wheels during turning -> Choose wheels and detachable hitch -> Trailing-arm suspension design, retractable -> Work with hull stress-analysis group -> Nacelle covers to minimize turbulence and friction -> Hull and panel nesting for secure road transport -> Testing
Aka deployment. The outriggers and associated solar arrays must retract to yield a maximum 8-foot beam in compressed mode, yet accommodate the severe heeling stress at the aka-vaka joint. The general F-27-style folding concept is established, but we need materials selection, mechanical design of the linkages, automated retraction, over-center hysteresis, stress analysis around the receivers, and detachable kayak hardware design. -> Model linkage assy and subject to testing stress -> Determine retraction via lines; select lock-downs and hardware -> What detachable connection to the amas? -> Design beams, with accommodation of solar arrays -> What dihedral? Adjustable? -> Build it!
Solar panel packaging, most likely on NidaCore polypropylene honeycomb substrates with grab rails, non-skid perimeter, center-folding design with suitably scaled web, thermal analysis of expansion coefficients that may stress the panels-substrate interface, cable routing between Tedlar PV sandwich and substrate, connector choices, electrical configuration, half-power operation when rafted, calculation of allowable solar cell deflection to determine required substrate thickness under worst-case impact loads, and so on. -> Calculate substrate thickness and acquire -> Design web, hinging, stowage scheme, and cable routing -> Do the layup, designing in serviceability -> Electrical tests (including individual panel profiling) -> Integration with beam assemblies & mechanical tests
Automatic sand-mold generation from CAD for complex shapes (such as cowlings), working with Dave Berkstresser and Autodesk as first users of this new technique now under development. -> Choose test project and talk to Dave…
Marine, Nautical, and Safety Issues
Radar System including display repackaging, antenna placement, interface with the GPS server, and possible export of video to the Mac or PC for charting overlay and export to Mission Control. -> Find out if Furuno radar video is exportable -> Physical repackaging for console -> Integrate with NMEA data stream -> Extract radar images -> Study microwave exposure and optimum mounting site
Desalinator and water manager, including sensors for levels, salinity, etc. Alert user to imminent failures. This includes the head cooling heat exchanger as on BEHEMOTH, with a working fluid sealed in the system and thermally coupled through the hull to ambient seawater. Backup desalination to augment the reverse osmosis system will require fabrication of a solar still. -> Acquire R-O system and integrate with tanks & through-hull -> Head cooling system and performance analysis -> Instrument complete water system -> Build solar still
Sail design, computer modeling of trimaran balance and performance, placement of appendages, rudder design, seaworthiness, trim calculations, rigging, hydrodynamics, possible front-rudder for fast-tacking and CLR-tuning, thruster placement, and more. All this will be in cooperation with our sponsors, Nelson/Marek yacht design (who did the Stars and Stripes catamaran) and Sobstad Sails. -> Research alternative sails including rotors, wings, and foils
Thruster design, including cavitation issues, depth, prop dimensions and shape, rudder integration, speed, and mechanical deployment by remote control from the helm. The plan is to have two thrusters, one at the stern of each kayak outrigger, allowing a high degree of maneuverability as well as individual kayak propulsion via pedals. -> Thruster selection -> Mechanical mounting and deployment -> Controller software -> JATO-override (work with controls group) -> Steering and rudder integration -> Kayak pedal system
Miscellaneous Subsystems, Gizmology, and pure EE
RFI/EMI reduction. The industry is desperate for experts in this field, and since computers and radios are mortal enemies by nature we can assume that it will be… well, interesting. This group will work with all controls and comm groups. -> Participate in cabling design -> Enclosure specs and EMI suppression -> PSD plots of noise leakage in all modes -> RF hot-spot minimization when transmitting
Antenna farm, including grounding, tuning, and performance analysis. This will be extremely challenging on a small craft. Copper mesh will be glassed into the hull as counterpoise coupling to the water… is this enough? How do the antennas interact, and what happens when the mast is down? -> Hull mesh design and cabling -> Lightning protection -> HF dipole, vertical, or backstay sloper? -> Placement of VHF, UHF, cellular, differential, and other antennas -> SWR and radiation pattern analysis in all modes -> Deployable cellular, VHF, and UHF yagis
Diagnostic Toolset. DPM, status bit multiplexer, coordination of test points, external network hooks, self-test facilities, and more. While testability is the responsibility of each control system group, centralized tools will require a separate project to ensure uniformity and consideration during cable harness design. This includes working with the power management group to map the 24 individual solar panel outputs to predict impending failure of any unit in the array. -> Design a self-scanning analog mux for DPM power monitor -> Clone the bike’s LED matrix bit-scanner -> Specify scope points to other groups -> Solar array mapping system -> Test loops in connectors for configuration self-check
Cable Harness. Far from being a boring passive component, this is the critical interconnect matrix for the entire craft. In this harsh environment, proper connector and wire choices will make the difference between success and failure of the whole system. We’ll have help from Scripps and MBARI, as well as a new sponsor that specializes in military cabling. -> Low-loss power distribution from solar array -> Battery bus standard -> Audio distribution standard -> Waterproof connectors -> RF shielding and crosstalk minimization -> Coaxial cable choices -> Through-bulkhead techniques -> Diagnostic hooks -> Multiplexing and fiber if appropriate
LED lighting system. As on BEHEMOTH, we’ll use high-brightness LED clusters everywhere except where white light is required. Properly potted with internal switching supplies, these have strong product potential in the marine marketplace, drawing 1/6 the power of standard products. -> Determine legal lighting requirements -> Design integrated bi-color and mono-color lights; pot clear -> Red LED lighting in chart table and console -> Spot and flood lights
Support Tasks, Logistics, and Management
Microship Project Partner. This is a big one. It has become quite clear that even with extensive engineering help and base-office management, I need a full-time project partner. This may be the person who occupies the ship’s crew module as well… so if you yearn for high-tech adventure, have extensive business experience, deal well with the media, express yourself clearly, feel comfortable living on the Net, and are free to devote the bulk of your time to a large-scale project (expenses paid), please contact me.
Engineering Management Team. A critical function during this development effort will be active, hands-on mangement of a large number of simultaneous projects. I am assembling a management team of three people, one each from ECE, ME, and CS — they will participate in ALL the projects to ensure consistency, good progress, availability of resources, inter-group communication, and access to help when needed. This group will provide me with frequent updates and alerts, and also help produce the more-or-less-daily project status report emailed to all participants and active sponsors. Industry is desperate for people who can manage projects WELL, and are not just MBA’s plopped atop an engineering group. If you’re on your way to an engineering management career path, talk with me about this.
Project Archive Management. Much of the support for these technomadic projects can be traced to an active program of publishing technical details… essentially sharing all the intellectual property related to the system in exchange for product donations, the assistance of wizards, and continuing media interest. This is fine, but the flip side is the time involved… I could easily spend all my time writing and none doing engineering. The project needs an archiver to handle video, still photographs for publication, status reports from workgroups, the CAD documentation binder, and help with monthly columns and other published material. This person or group has to be literate, precise, and willing to meet hard deadlines.
FTP and Listserv Maintenance. Closely related to the foregoing, this involves keeping the electronic archives and mailing lists current. Nearly 10,000 people receive project updates via the Internet, and in my chronic state of overload, reports are far less timely than they should be. While they should still carry my writing style, some help with organization, deadline reminders, list updates, and ftp site maintenance would be extremely useful.
Database Maintenance. Help keep up with databases of contacts, sponsors, resources, and ship inventory. This includes batch geocoding of all addresses to allow integration with GeoQuery. There will also be a file of all on-board components, detailing weight, VCG, LCG, cost, spares, vendor, and so on — we need someone to manage these files and make sure they are well backed up.
Admistrative and Support Tasks. We need a lot of help in this area: photocopies, errands, document generation, filing, sending for literature, purchasing, and general office management. If you’re heading for a career in administration, this will give you good experience (and I’m willing to pay for help with a few particularly boring tasks that nobody in their right mind should volunteer to do for free!).
o o o o o o o o o
Most of the projects can be developed by small teams, or in some cases individuals. Yet they are all connected, which enforces system awareness and standardization of protocols, uniform packaging, consistent documentation, and so on. This is very good discipline for the engineering profession in general. Despite the serious aspects, however, we must never forget the constant undercurrent of fun… which is, after all, the bottom line. Also, this is largely supported by sponsors, so there will be some very interesting new toys for us to play with (OK, I confess: this is part of the addiction that has kept my high-tech nomadness alive for ten years).
I look forward to your comments and suggestions on any aspect of this. And again, remember that the project needs help at all levels… before this is over, we’ll have a spirited team of dozens, earnestly working to get me out of town. Please contact me with your interests!