I am lucky enough to have a friend that will take me sailing, and now that I have a focus on data visualization I realized there is a lot of information you need for sailing that you can't see or quantify completely on your own. Wind speed and direction (obviously), depth, obstacles, currents, speed, landmarks, coastlines, weather patterns, channels, position, orientation, you need to visualize the information around you and it needs to be done so that you can plan your route and act in the moment when plans are thwarted by new developments. The data involved in sailing is something we all can understand, and it's something that been developed for thousands of years, but in looking at the wind vane on the mast, the wind ribbons on the sail, and the flag on the mainsail rigging, I couldn't help but thing about how to capture all this information and optimize our sailing algorithm. Wanting to digitize sailing information isn't new, but sailing brings to Earth our esoteric ideas and provides us data scientists with clear purpose in how we visualize our data.
Collecting & Displaying Sailing Data
When sailing, it's pretty clear when you're doing something wrong: you're not moving, the sails are flapping, you're heading towards a rock, that sort of thing. However, once you're moving it's very hard to tell whether slight changes are really making a difference, and once I had the head-sail ribbons catching the wind, more or less, I wasn't sure how trimming the sail or changing the angle of the main sail made any difference. My first instinct was to put sensors all over the boat so I could monitor the hull, the sails, the rigging, the lines, everything. I could know just how fast the wind was transferring from the head sail to the main sail, how big a vortex was created, whether the sail was pulling along the axis of the boat, whether I was capturing evenly along the length of the sail. Everything. My second thought, as we prepared to tack, was how would I use this data, how would I make sense of it in real time?
I realized some sensors already existed, in the flags, ribbons, and vanes conveying wind information, in a way they were augmented reality. Of course I took it to the next level and imagined something like Google Glasses showing laminar flow of the wind across the sails, with color representing the speed. I imagined similar with the mast, hull, rigging, and lines, but in their cases color represented pressure magnitude.
I didn't just want to visualize the raw data, I wanted some indication of what was 'best', as if were were in some over-determined system with a closed form answer that could be found. Perhaps a calculation showing the total pressure on the sail, or the effective pull of the combination of both sails, in any case I could tell I was building an artificial intelligence system for robotic sailing. The difference between sailing and robotic sailing is subjective versus objective respectively, with subjective including people as part of the real-time decision processes. So what is the balance between abstracting the raw data and still being part of the sailing process?
As always happens when your imagination triggers new searches, you usually find someone is your allay and is walking down a similar road. In 2010 BMW and Oracle won the America's Cup with a fixed wing sail design, trimaran, hydrofoils, and a number of other innovations including wireless sensors throughout the boat relaying real-time information to wrist PDA's customized for each skipper's roles and duties. I knew about the Oracle racing team that won the America's cup in 2010, which featured fixed wing sails, but now that I was thinking about data I found out just how extensive their use of real time data was. While I don't think I'll be able to afford the heads-up display they used in addition to their wrist PDA's, there are a number of wireless sail sensors you can buy on the market, and my friends birthday isn't far away...