A revolution in routing

A 388
In simplistic terms, the winning combination in ocean racing has always been speed, strategy and disaster avoidance. Boat speed – typically achieved through incremental gear refinement and superior crew work – is particularly critical in short-course and medium-distance racing. By contrast, when a race continues over days or weeks and covers thousands of ocean miles, strategic decision-making often becomes increasingly, even dominantly, important. Sailing a near-optimal route has often been the key to victory, but finding this ideal course is rarely a simple task. The most important environmental factors – wind, current and sea surface conditions – are complex and unstable, with ties to almost every aspect of physical oceanography and atmospheric science.New role for navigators

Thanks to the near-universal use of GPS in modern ocean racing, position, track and speed over ground now fall under the category of raw data. These days, ocean racing navigators concern themselves almost entirely with routing - the task of evaluating course options in the context of boat performance and the sailing environment. Amassing accurate, timely weather information is, of course, key to routing success. The advent of offshore Internet access opens a treasure trove of potentially valuable information (albeit at fairly high cost when it involves slowly downloading big files at satellite airtime rates). But in high-budget events like the ongoing Volvo Ocean Race (VOR), navigators need to guard against being buried in a sea of information - everything from real-time satellite photos in the visual, infrared and microwave spectra to meteorological analyses generated by hundreds of sources from around the world.

With unfettered phone and data communications now available virtually anywhere on earth, the traditional prohibitions against outside assistance during sail races have lately come under scrutiny. At one extreme, a few events now allow shore-based routing experts to communicate strategy to a crew while racing is underway. At the other extreme, there are races that severely curtail the use of electronic aids to encourage traditional skills, such as celestial navigation, and to some extent, reduce the cost of competition. Most often, however, race organizers settle somewhere within a broad middle-ground, where shore-side routing assistance is excluded, but access to professional weather analysis from various public domain sources is permitted, along with the use of onboard computers running routing software. Specific details of what is or is not allowed can have a huge impact upon the character of a particular event. This is neatly illustrated by two recent events, last fall's Mini Transat and the ongoing VOR.At opposite extremes

The Transat 650 Charente-Maritime Bahia (still popularly known as the Mini Transat) is a solo endurance race sailed in powerful, dinghy-like 22-footers. The 2001 race covered a 4,000-nm course in two stages, starting in France and finishing in Brazil, with the scoring based on cumulative time.

Among the most demanding of all ocean races, the Mini Transat regulations pretty much guarantee that competitors will be on their own at sea by discouraging non-emergency communications, as well as most electronic navigational devices. Mini sailors are allowed position-only GPS (no plotting functions), basic calculators and high-performance autopilots. They can access VHF and SSB voice forecasts but must agree not to receive routing advice or other individualized outside assistance while racing. An Argos beacon aboard each boat enables race organizers, safety vessels and Web spectators to track the fleet; but except when other racers are in sight or radio news broadcasts mention the leaders, competitors seldom have more than a vague idea of their fleet standing, let alone the actual geographic positions of the competition.

The Mini-class circuit attracts an eclectic mix of adventurers, seriously competitive amateurs, and professional sailors looking to broaden their resumes and attract the interest of major sponsors. The Mini Transat itself is an inherently dangerous race that has suffered a number of fatalities over the years. Despite extensive safety precautions, there was another life lost in 2001 - Italian sailor Roberto Varinelli, who fell overboard during Leg 1.

Times are cumulative for the two legs, so those competitors who suffered significant damage in the stormy upwind conditions encountered in the Bay of Biscay (a common occurrence), were essentially out of the running for overall prizes. One of these was British sailor Brian Thompson, a top international pro in crewed racing, who had been considered a pre-start favorite. Deeply dissatisfied with his seventh-place showing on Leg 1, he led the fleet nearly all the way in Leg 2, only to be pipped at the line by overall winner, Frenchman, Yannick Bestaven. Interestingly, Bestaven passed Thompson thanks to pre-race routing assistance - local knowledge gleaned from Brazilian sailors who advised an inshore approach among the fishnets for more breeze and a hotter reaching angle into the finish.

In Thompson's view, the lack of outside information, particularly competitors' positions, is a defining feature of the Mini Transat. While approaching the doldrums (aiming to cross at the longitude he'd pre-selected with the assistance of his pre-start routing specialist), Thompson believed he was positioned on the eastern edge of the fleet. In reality, he was the farthest west. By the same token, Bestaven crossed the line believing that Thompson had long since finished.

A race format that essentially forces competitor's to do their own thing, with little opportunity to respond tactically to the competition, may well be a benefit to competitors with less experience or fewer resources than the pros. In 2001, second place overall went to Simon Curwen, another Brit, but a talented 42-year-old amateur who'd taken a year's sabbatical from his environmental consulting business for a go at this particular race. A former collegiate and recreational racer, Curwen had purchased pro sailor Lionel Lemonchois' beautifully prepared Mini after the 1999 Transat and put in the necessary training time to really learn the boat. During both legs of the race, he sailed efficiently and conservatively, rarely falling out below fifth in the fleet rankings. He estimates that his yearlong campaign will have cost about $100,000 (after resale of the boat), and having enjoyed his mid-life crisis, he has no firm plans to continue in solo racing.

In sharp contrast to Mini racing, where a successful sailor must be skipper, navigator, tactician and deck hand all rolled into one, the VOR is increasingly a game for specialists. The VOR format encourages this specialization by permitting unlimited crew substitutions before each new leg and by allowing extensive access to the Internet for gathering weather and environmental information. In addition, all the teams employ full-time weather-routing experts who bone up on the upcoming legs while the racing is underway and work intensively with the navigators and skippers during layovers.

Each of the eight VOR boats has at least one full-time navigator (and most would quite likely have engaged two if enough qualified personnel had been available). The navigator's workload is staggering, and working conditions are routinely horrendous. In some respects, access to the World Wide Web has created a monster for the VOR navigators.

Prior to the start, each crew was permitted to designate 10 websites that both they and the other teams could access freely during the race. To ensure no other sites get visited, all e-traffic from the VOR yachts is routed through special communications equipment at race headquarters.

In principle, there could have been 80 available sites, but with duplication, the final tally came to 47 (see www.volvooceanrace.org/news/tactics/prerace/t0_010921_websites.html). Since quite a few of these designated sites are the principle Web addresses for sprawling governmental organizations, which include links to numerous subsidiary sites, 47 is probably more than enough. Inmarsat B and Mini M communications are still quite expensive, and selecting freely from the wealth of available GRIB files, satellite images, etc., can easily lead to round-the-clock downloading. Some estimates of individual team satcom budgets have run as high as $50,000.

Whether all the nominated sites offer useful information is a matter of conjecture. Several VOR navigators have suggested that some selections might be red herrings, tossed out so rival teams would squander time and resources attempting to determine their significance!

At press time, the VOR boats had arrived at Rio de Janeiro, having sailed about three-quarters of the overall distance, yet having completed only four of the nine legs. So far, the new points-based format has done an admirable job of keeping the race tight and exciting - far closer than would have been the case under the old cumulative time scoring. After four legs, the meticulously prepared illbruck Challenge is seven points ahead on the leader board, (as many had expected), but there's a very tight fight underway for second with Grant Dalton's Amer Sports One only two points ahead of Assa Abloy and four other teams lined up right behind, each separated by just a single point.

With the overall scoring this tight, almost anything could happen. For example, Assa Abloy, navigated by the previous race winner Mark Rudiger, came roaring back from a devastating strategic misstep on Leg 1 and currently is in third place on the score sheet. Leg 3 was a huge boost for this team - line honors in the Sydney-Hobart Race and first boat into Auckland.

Interestingly, despite the massive effort that every team has put into weather strategy, this marathon is shaping up to be a more tactical contest than the last VOR. By and large, the boats are being sailed very much as though they were short-course racing - covering the opposition and using weather analysis primarily to seek small gains or minimize the risk of getting passed. This, of course, is another consequence of the new scoring format in which it makes no difference whether you're a boatlength ahead of an opponent at the finish or 100 miles.Onboard routing solutions

Working with extensive raw weather data and public domain predictions as downloaded from the Internet, VOR navigators typically spend hours a day evaluating various what-if scenarios with the aid of sophisticated routing programs. Because they know the positions of their competitors (updated every six hours), they also have the option of modeling routing for another boat, partly to determine whether the separation is likely to confer an advantage and partly to predict what the opponent might try next. It's a computer game of enormous complexity, but one that's rooted in real-world racing experience. Top navigators must be technically savvy individuals who can keep the hardware running, keep their minds on the big picture, and avoid falling prey to information overload.

Currently, the two biggest names in routing software - RayTech and MaxSea - are both being used in the VOR. Raymarine developed RayTech Sail Racer through its acquisition of Kiwitech, the well-known New Zealand firm that began doing pioneering work in computerized routing back in the 1980s. MaxSea, a product of the French company Informatique & Mer, also got its start about a decade ago and to date has been the software most widely used by marathon sailors. Lately, MaxSea has also expanded into the tactical racing arena through the development of MaxSea Racer. There's also SailMath, developer of the popular Deckman programs.

Both MaxSea (www.maxsea.com) and RayTech (www.raytechonline.com) construct optimized routes by creating a sequential series of isochrones. Each isochrone is the chronological equivalent of a contour line. It links up all the various points that a fleet of identical boats would theoretically reach at the end of a selected time interval - say six hours - after fanning out from a single starting point on different points of sail. Subsequent isochrones are derived by treating numerous loci on each preceding isochrone as the new starting points for an equal number of hypothetical fleets. As the route map evolves, the optimal route ordinarily emerges as a bulge in the outermost isochrones.

A relative newcomer, California software developer David Brayshaw (www.goflow.com) has taken a somewhat different approach to routing that employs a fine-scale digital grid covering the sailing area. The times required for a yacht to sail between adjoining points on the grid (either directly or by tacking and gybing) are computed individually in view of the wind and current conditions applicable to that part of the course. These times are then added together in various combinations to construct and evaluate routing alternatives.

Most routing programs derive a routing solution based on a snapshot of wind conditions as shown on a single, digitized forecast or GRIB file. Naturally, during an extended passage, the wind conditions will often change dramatically while the yacht is at sea - possibly reversing the advantage of a route that had initially looked the most promising. In large part, the art of routing (as opposed to the computer science) consists of developing a strategy that will minimize potential setbacks if the weather evolves differently than expected. In addition to running multiple routing solutions based on differing short- and long-term forecasts, offshore racing navigators often draw heavily on their pre-race analyses of historical weather patterns. For the VOR, these background weather studies may entail an effort best measured in man-years.

But by the same token, it's worth remembering that the typical voyaging sailor is probably more concerned about avoiding strong headwinds or stormy conditions during the upcoming few days than in trimming a day or two off an extended passage. For this scenario, the routing program offers a straightforward solution because it can readily be used to produce an updated route every time an updated forecast arrives. And when two alternative routes are not all that far apart, in terms of projected passage times, one will often be favored due to a lower risk of adverse weather or reduced exposure to other hazards.

It's also worth noting that even with good weather information downloaded from the net or via SSB, the value of a routing program will be degraded significantly if the polar data for the yacht is not up to snuff. Obviously, this is not an issue for the VOR teams, which spend months testing equipment and gathering precise performance data for every conceivable sail combination, apparent wind angle and sea state. However, a voyaging sailor might not be keen to spend big bucks on high-end wind instrumentation or the time required to construct a personalized polar diagram. Stock boat polars (sometimes available from designers or through U.S. Sailing for International Measurement System-rated sisterships) are the next-best thing.Corinthian ideals and legal realities

How much routing assistance is too much? The debate continues to escalate as access to professional weather analyses and onboard computing power grows by leaps and bounds. A few events like the Race and the Jules Verne Trophy allow unrestricted participation by shore-based routers, but this remains a rarity. Most sailors and race organizers still apparently believe that bluewater racers should be largely self-reliant, and that strategic decision-making is properly left to the sailors onboard. Coaching before the start has become an integral aspect of the modern racing game in everything from Olympic dinghies to sizeable offshore one-designs, but to allow the coach to participate directly via telecommunications is, for many, a step across the proverbial line.

Things get trickier, however, when Corinthian racing principles butt up against another esteemed maritime value: doing everything possible to ensure safety at sea. In explaining why they decided to allow almost unrestricted onboard Internet browsing, VOR organizers pointed out that to deny access to a technology that could enhance marine safety would be morally and legally unsupportable.

Obviously, this is a view that the marine insurance industry is sure to endorse and something that promoters of purist events, such as the Mini Transat, will ultimately need to reconcile with their philosophical stances.


Contributing Editor Sven Donaldson is a writer and former sailmaker who lives in Vancouver, British Columbia.

By Ocean Navigator