A Sampler of On-Board Applications

It occurs to me as I near the end of this lengthy tale about the Microship project that the massive amount of on-board technology may, at first glance, seem a bit excessive. I mean, even with the luxury of a graphic front end, there’s only so much low-level control that you can do on what is basically a canoe with training wheels. Beyond the obvious marine navigation and fundamental technomadic needs, what might we want to actually DO with all this?

So before we leave our little peek under the hood, I’d like to give you a quick look at some of our other applications for this boatload o’ goodies...

System Diagnostics

One of the original motives for the whole data display system is simply being able to keep track of what’s going on in the system itself. One could argue that by eliminating the system, I would also eliminate the need to keep an eye on the system... thus knocking a year or two off the project and cutting weight by a hundred pounds or so. But a big part of this is the underlying geek expressionism that requires a dose of complexity in the midst of languid escape from the madness of civilization.

The physical Microship components generally don’t require much in the way of instrumentation – I can clearly see and feel them, and on a boat this small, visceral seat-of-the pants feedback is probably a better indicator of sail trim than, say, microbending-lossy optical fibers embedded in the mast and strain gages on the blocks to report stress levels. But there are lots of things I can’t see that are just as critical.

Take the solar panels, for example – a vast array for such a small craft, totaling sixteen 30-watt modules bonded to eight vacuum-bagged foam-core panels, hinged into four assemblies that fill the spaces between the hulls. With the peak power trackers extracting some 15% performance improvement, we’re expecting as much as 36 amps (at 12V) under ideal conditions. But all sorts of things will reduce this ideal number: dirt, insolation angle, elevated temperature, shading, saltwater incursion, wiring failures, and tracker demise. As such, the eight channels of 60 watts each are continuously reporting their performance to the server, viewable as real-time stripcharts on the Microship solar array web page... little Java sandboxes that fill with data from the database (depending on selected range), then scroll new data into the graphs from right to left. This not only lets me view them all in a stack to observe a possible problem, but also explore historical performance and extract statistical information... as can the power-management guru who designed the peak power trackers, logging in from his home in Wisconsin.
?This same idea extends to other subsystems: console pressurization and sealing, thruster control, humidity inside hatches, bilge pump cycling, internal power distribution, system reboots, and so on. Being able to collect and manipulate data makes a wide variety of diagnostic tools fairly trivial to implement.


Environmental Data Collection

In a rather similar vein, from the perspective of the systems involved, we are taking advantage of on-board technology to do our small part to help clean up polluted waters. With years of anticipated coastal/inland on-water time and the technological tools on hand to collect and transmit environmental data, it would seem almost negligent not to do so!

We have no particular political or scientific agenda, and are thus gathering a wide range of easy-to-capture telemetry data. Some of it is purely internal as noted above, and some is navigational (speed, GPS coordinates, heading, and the like). But quite a bit falls into the broad category of environmental measurement:

• Air temperature
• Ambient light
• Barometric Pressure
• Dissolved oxygen
• Humidity
• Ozone
• Radiation (both air and water)
• Ultraviolet intensity
• Water pH
• Water salinity
• Water temperature
• Water turbidity
• Wind direction
• Wind speed

We are seeking other sensors that easily yield relevant and interesting data, with particular interest in hydrocarbons and ion-specific indicators of common industrial and agricultural pollutants (we’re seriously lusting after a Voyager portable Gas Chromatograph from Perkin-Elmer). Although we’re not doing “real science” in the formal transect ‘n survey sense, we do have the opportunity to collect location-stamped telemetry blocks at any interval and transmit them in real-time to a publicly-accessible web browser. The resulting river profiles should yield interesting data as we pass industrial areas, population centers, and regions of heavy agricultural runoff... and to make it easy to use, we are building a server-resident database to host the raw data along with a set of front-end tools that allow graphic observation of any group of channels over time – including the ability to click anywhere on our “breadcrumb trail” for a snapshot of the values at that point accompanied by corresponding observational text and images.

Graphic presentation is the first step in making raw data interesting, and – with the added rich context of the Microship adventure – we are targeting this system toward schools. To really deliver something of value (one-pass data is too much affected by rains and seasonal variations to allow more than casual conclusions except in obvious point-source pollution cases), we must go beyond the data itself and port the entire set of hardware and software tools to our educational clients. Students who grow enamored of nomadic data collection by watching graphic sensor logs of our cruise downriver can copy the designs, download the software, and apply the same techniques to local environmental problems. Such continuous monitoring over an extended period can indeed yield real science, not to mention a firm grasp of principles ranging from biology to wireless networking. These designs will allow relatively non-technical folks to build unattended monitoring systems that communicate with classroom computers via a variety of data links, inviting a whole set of potentially interesting student projects and collaboration with the scientific community.

Along these lines, we might present the Microships to schools as a sort of “experiment bus” along the lines of the Space Shuttle (but much more likely to drop in for a classroom visit!), perhaps with a competition that would encourage students to design a new experiment or sensor channel that we could add to the system. Coupled with a series of on-site school appearances while underway, this should make the whole affair quite engaging as an educational project; during the BEHEMOTH era, I delighted in getting kids turned on to technology through something “cool”... a much-needed alternative to dry textbooks.

Communications

This is a pretty broad topic, but justifies a bit of commentary since it affects just about everything else on board. Much of what we are talking about depends on a variety of local and global voice and data links, and on a vessel this size it becomes a real problem to package all the radios and antennas without creating a quagmire of interference problems. Processor boards spewing broadband RFI are enough of a problem...

Let’s take a tally of the radio links we need. On the local level, there are a number of communication paths connecting Io and Europa with each other and the backpacks: wireless ethernet for high-speed data and a suite of IP services, the “Packnet” that lets low-level guys chat without involving the big iron, amateur dual-band HT’s and license-free FRS walkie-talkies for casual person-to-person chitchat, and high-quality bidirectional video and audio links to extend the crossbar networks across both boatlets with minimal insertion loss. All these are relatively short range links, and at the rate things are going, we might find that they all become virtual circuits through a single high-speed spread-spectrum data link. (The pace of technological change is why we wait until the last minute to acquire computers, wireless data hardware, and nav tools.)

When we start to look outside our little flotilla, a new set of tools is necessary, although there is some overlap. A key voice accessory is a handheld dual-band ham radio transceiver for line-of-sight communication with locals and access to repeaters, and the data equivalent is a minor reconfiguration of the existing packet radio link to connect with local gateways. This is the simplest tool for our telemetry transmission – just drop email into the amateur radio store-and-forward packet network. In the boonies, with no local repeaters or digipeaters, we’ll switch to HF PACTOR... as well as APRS, the position-reporting protocol layer atop amateur packet radio.

This is neat stuff. APRS appears in the system as a small box that automatically transmits position reports between our boats and packs at any preset interval. But it’s not just local – a worldwide network on 144.39 and 10.151 MHz collects and forwards all such transmissions to a central Internet-delivered data stream, allowing our location at any moment to be viewed on a map through any web browser. It’s all quite magical – have a look at http://map.aprs.net/microship, and if we have transmitted a position report in the past 8 hours you’ll see us with 500 ft precision in three successively zoomed MapBlast windows. Ahhh, technology...

APRS offers some capability for additional telemetry channels, and we’ll transmit a subset of our data stream through this path. But the full blocks will consist of 50 or so channels, and will be accompanied by current text reports, embedded notes for our home base, and still-frame image data. One of the linux server’s jobs is to assemble these automatically generated bits of email, then choose the most appropriate path based on cost, location, and content. The choices at the moment are amateur packet via VHF or HF, cellular modem, or Globalstar low-earth orbit satellites. The latter is a fully operational worldwide commercial network, and is expected to become the backbone of our wide-area network connection with the rest of the world.

As wireless network infrastructure grows, we’ll be experimenting with spread-spectrum radios operating under both FCC part 15 (license free) and part 95 (ham radio, with business traffic prohibited). This technology is in a state of flux, but our goal is clear: maximum bandwidth at the lowest financial and power cost, regardless of location. We’ll be depending on this not only for the telemetry and email, but also to support remote software development and system administration, security when we’re away from the Microships, server backups for essential files, website maintenance, and everything else that has moved Internet access squarely into the middle of Maslow’s hierarchy of needs.


Video Production

One of the more whizzy and exciting components of the Microship project is something that I always wanted with BEHEMOTH: video. Lots of video. Cameras and hydrophones underwater, cameras on deck. Cameras in the consoles, staring at our faces. Even cameras that can be remotely steered from afar...

The most amusing video source, in fact, is a video turret on Europa’s arch, operating under software control as a node on Natasha’s Shipnet. Two cameras – a high-quality zoomable color unit and an infrared/BW low-light microvideo unit – are mounted on a rotating platform that can be quickly directed at any angle, or set to scan at any speed between any two angles. This has obvious value as part of the security system, and can unobtrusively gaze at any scene of interest, even under the control of a web browser thousands of miles away. On Io, similar functionality with somewhat lower video quality is accomplished by adding a small color/BW/IR/laser pod atop an az-el remote-controlled searchlight with added position feedback.

The Vixbar (video crossbar) integrates all these sources across both boats, both consoles carry monitors, and a software-controlled 8mm video cassette recorder can record any channel. There’s even a video transmitter that can send any requested channel to a pocket TV, though we may toss that now that wireless bandwidth is getting to the point that we can just pipe compressed video streams to PCMCIA frame grabbers in our backpack laptops.

But the sexiest part of all this is the production system – a nonlinear digital video editor known as the Draco Casablanca. This is integrated into Natasha’s console, allowing her to work from any source material (including digital video cameras and computer-generated images), manipulate clips with frame-precise cuts, assemble a production with a storyboard, generate hundreds of different transitions and titling effects, mix voice-over and music tracks, and output the whole finished work with broadcast quality... all from the confines of a canoe hull. An internal Quantum 18 Gigabyte hard drive can store nearly an hour and a half of professional-quality video, and these are easily swappable to allow her to work on multiple projects.

While we travel, we’ll produce video tapes of our adventures, the technology that makes it all possible, and amusing encounters along the way – a major upgrade to the text-only storytelling that has always supported my technomadic adventures.

Stay tuned!


Music

While we’re on the audiovisual subject, I should mention another key application that is being designed into the Microship system. Actually, it’s two applications...

First, it goes without saying that a decent stereo and music library is essential, and it’s amusing to recall just how close I got to buying a Sony 150-CD changer and interfacing it to provide a mobile music library. I’m sure glad this whole gig is coming together at the turn of the century, for now we have MP3 instead – the music compression algorithm that has swept the Internet and caused the recording industry more than a little panic. With roughly 10:1 compression ratios, CD-quality audio can occupy about one megabyte of disk space per minute of playing time... or 16 hours per gigabyte. At the vanishing cost of hard drives these days, there’s no longer any reason to carry those bulky CD’s that were such a key resource on BEHEMOTH.

In fact, experiments here have demonstrated that only one of us needs to tie up the disk space, since streaming a megabyte/minute via wireless ethernet is trivial. The central linux jukebox can easily support two users listening to different material simultaneously while going about its other jobs unperturbed.
And speaking of high-effiency music storage, there’s another system on board that should offer endless amusement... a full-scale music studio consisting of a MIDI synthesizer module, drum machine, sequencer software, multi-effects processor, mixer, and recording tools. This little ensemble will not only play the virtually infinite variety of .mid files on the Net, but also provides a computer-generated backup band to make my improvisational jazz and blues flute playing more interesting (I get in a rut when playing solo). Naturally, all these components are interconnected via the crossbar networks, so any resource combination is only a quick script command away... like generating music tracks for video production, adding amusing effects to the local response to security violations, and having endlessly interconnectable toys to while away the hours of a long crossing.


Security

Finally, I should say a few words about one of the more compelling “high-level” applications – keeping the boatlets safe and, equally important, reducing our paranoia index in a wide variety of situations. The best lock is the human eye, but I know from long bike experience that being tied to the machines when there other interesting things to do gets boring.

In my travels with BEHEMOTH, it didn’t take long to discover the twin effects of a high-tech intriguing contraption that draws a crowd and enough interactivity to convince anyone of dubious motives that they were dealing with something mysteriously intelligent and connected to larger forces Out There. The crowd-pleasing effect makes it impossible for someone to anonymously toss it in the back of a pickup and drive off; people would notice and most likely stand around asking questions. And the interactivity was amazingly effective... when the bike responded to touch by intoning “Do not touch, or you will be vaporized by a laser beam,” the effect in some cultures was dramatic, everybody backin’ off and lookin’ at the bike real s’picious-like. Of course, when it said the same thing one day in Apple’s cafeteria, it just caused engineers to swarm all over it while buzzing excitedly about the Berkeley Speech Technologies algorithm and whether they had been detected by microwave Doppler or flux-density variations around a puddle of mercury... but I wasn’t too worried about those guys haulin’ the bike off to a garage and partin’ her out.

In a more serious vein, it was comforting to know that the bike would trigger my pocket beeper in response to a security violation, and even do the 911 “I’m being stolen and my coordinates are...” call if the worst happened.

On the Microships, we’re taking a similar approach but with more subtlety and bandwidth. If we’re nearby and any of a variety of security sensors are triggered, then all sorts of options are available including automatically scanning the area with the video turret and recording or transmitting live footage of the perpetrators, beaconing emergency help requests via packet or satellite, carrying on a running and vaguely threatening commentary with the speech synthesizer, generating loud noises, flashing lights, phoning the Coast Guard or hollering on marine VHF, and unleashing a few nasties. The choice depends on a simple setup tool which allows us to put checkmarks in the intersections of a large matrix of horizontal sensor lines and vertical response lines, then storing various configurations as macros for different environments (marina slips, deserted beaches, campgrounds, urban moorings, friends’ garages, and locked corporate meeting rooms all call for different security system configurations).

The nice thing about this, as with most of the other applications, is that we’ve put so much work into the infrastructure of crossbars and networks that it’s trivial to accomplish the rerouting of resources with a few lines of code or mouse clicks on a browser window. The net effect, so to speak, should be a very amusing little pair of boats...

With that, let’s wrap this up with a quick look at project management!