Putting the “Spin” on a Robotic Sailboat

When it’s finished, Christopher Miller’s modified 2.4mR class keelboat will be able to sail itself. “Herb’s Watch” for May 7, 2009

robo boat 368

Lin Pardey

Christopher Miller has a bit of time on his hands. But he’s found himself a little project that’s certainly filling up the hours. Originally from Wisconsin, where he grew up racing Lightnings, by profession he’s a software developer and engineer whose resume includes stints at such powerhouses as Apple and AOL. A few years ago, he bought a 52-foot cruising boat and struck out across the Pacific, eventually landing in New Zealand, where he bought an island home, started a family, and talked himself into a gig that combined his two passions‚sailing and technology‚in the sail-development program at Emirates Team New Zealand. The job with ETNZ was going great until, you know, the America’s Cup went totally to hell and came to a crashing halt.

But Miller isn’t exactly the sort of guy to sit around staring at the clouds. Instead, he bought himself a modified 2.4mR class keelboat, a one-man vessel commonly used by disabled sailors, that he calls Spin. His neighbors in New Zealand, the sailing writers and voyagers Lin and Larry Pardey, are old pals of mine who thought I might be interested in what he was up to. Over the last few weeks, as Miller’s continued to develop the concept and begun tests and sea trials, we’ve swapped several emails about the project. So, what, exactly is Spin? Let’s let him explain:

“Spin serves a variety of purposes. Primarily she’s a test vehicle for the commercial sail-vision system I’m developing. Her size enables me to do everything single-handed: stepping the rig, working at the masthead, launching, sailing, transporting, etc. She’s relatively economical to modify and equip due to her petite size but large enough for me to sail and fit out with full-size nav instruments. My sail-vision hardware and software can operate on Spin just as well as (the IACC yacht) NZL 92 because she has similarly shaped and marked sails. Spin and her rig can be stowed safely and neatly into a 20-foot container.”


So it’s clear that Spin, at least initially, was conceived as a vehicle to continue and build upon Miller’s America’s Cup research. The problem he hoped to address was, paradoxically, simple and complex. All AC teams still struggle with inaccurate true wind direction (TWD) due to the cumulative errors in its precursors: boatspeed, heading, leeway, mast twist, upwash, and a few others, he explained.

Inaccurate TWD implies inaccurate true wind angle (TWA) which implies inaccurate VMG, the very thing that sailors are trying to optimize! On an AC yacht the TWD isn’t even accurate within 30 degrees when she is dialed up, standing still, and the navigator is trying to figure out the line bias based upon TWD. For me to make headway on the TWD problem I need complete control over a boat to determine how the variables interact, which means better sensors and measurement of things that are not normally measured on a yacht. This means better boat speed, better compass calibrations, real-time sail shape, better water and wind sensors, control of sheeting on, easing off, turning, etc. Hence the servo-winches and actuators that offer virtually complete control over line position and load.

Once Miller got into the project, however, he realized its scope and potential had benefits beyond the America’s Cup arena. Entering Spin in robotic sailing competitions, and using what he was learning to help disabled sailors reach and exceed their on-the-water enjoyment and possibilities, also became very real parts of the equation.


In mid-March, Miller was just about to take Spin for a, well, spin, and sent the following update: “Spin should be going for her first servo-controlled sail later this week. Spin’s mainsheet, jib sheets, jib furler, runners, boom vang, and helm have already been servo’d. The main furler, halyard, and traveler will be servo’d after the current gear has been tested. Then the real fun can begin towards the long-term goal: making Spin teach itself how to sail and adapt to conditions.

I also decided to take a few extra steps by making Spin robotic. Another possible long-term goal is to create an autonomous sailboat that is energy self-sufficient, can adapt to conditions, can operate for weeks or months at a time, and can carry a 50 kg. payload.”

A couple weeks later, Miller registered another important milestone and filed another update.


“Spin sailed under full servo-control of mainsail and jib today. This achievement has taken over a thousand man-hours to achieve…from a boat that was perfectly usable last year but fully manually controlled. I was onboard pushing joysticks & buttons making sure everything went smoothly. Sailing by digital joystick takes getting used to, especially in the tight maneuvering space of the Auckland Viaduct. Spin can tack in 3 seconds. The only technical issue was an unpleasant noise from the auxiliary motor/hydroelectric generator when pushed to 20 amps output. Next time I won’t scrimp on that critical component.

There are many months of work to do before she’s autonomous and net-energy-positive, but it’s a good start. The Sailvision worked from initial installation last year. Next projects are waterproofing, mainsail furler, instrumentation, and telemetry. This afternoon it’s time for a few rums and Foo Fighters in The Spin Lab.”

With no end in sight regarding the current America’s Cup imbroglio, Miller reckons his work on Spin will continue unabated. He plans on lightening the keel and fashioning a bolt-on extension to address increased righting moment with the addition of a new, untried, square-top mainsail. Participation in the 2010 World Robotic Sailing Championship is a strong possibility. Fund-raising is something that needs attention‚ “I’d love to find a sponsor like Parallax, the company that makes the microcontroller chips that Spin depends on”‚ but otherwise, the work continues.


I have to confess that much of what Miller has explained to me has, quite frankly, totally escaped my complete comprehension. I guess maybe he was thinking of me when he offered the following: ‚”The name Spin works really well on many different levels for this project. Spin also stands for Science Project for Instrumentation and Navigation. I can also say to people, “Go ahead, take her for a Spin” or “This crazy project makes my head Spin!”

That I understand.

Here is what Spin has and does:

  • Four onboard Parallax Propeller microcontrollers perform sail control, system monitoring, and navigation
  • Each Propeller resides on a custom Spin Controller PCB designed by Miller
  • Each Spin Controller supervises up to four motors/actuators and has 16 12-bit A/D channels
  • Sail control lines (sheets) and rudder are controlled using 7 gear motors and 3 linear actuators (so far)
  • Each motor has a dedicated motor controller receiving packetised serial commands from a Spin Controller
  • Each winch’s position, velocity, current, and motor temperature is measured at 30 Hz.
  • Motor end limits, home position, and max current stored in EEPROM
  • Spin Controllers simultaneously accept manual joystick, R/C PWM, and telemetry commands
  • Auxiliary propulsion is provided by a 600-watt trolling motor that also generates power while sailing
  • Navigation system includes AHRS (heading, heel, pitch), GPS, four boat speed sensors, and three anemometers
  • Communication system includes 900 MHz long-range telemetry and 2.4 GHz Wi-Fi.
  • Spin also has onboard cameras that measure sail shape in near-real time (a.k.a. Sailvision)
  • Spin is carried on a motorised trolley with dual 24V, 60A gear motors. Trolley can pull 300 kgs horizontally for going up gangways from floating pontoons.