Spotting the starting line by eye is a practice that has remained unchanged for generations. For small fleets, a simple look down the line works well, but as the line gets stretched out to accommodate a big fleet, calling the line by sight is an impossible task for even the most seasoned and eagle-eyed spotter. And bigger fleets are usually more aggressive, which compounds the problem. General recalls are no fun, unless you like starting practice, and black flags are real buzz killers.
Rumor has it that the black flag was invented on the fly with a skull and crossbones flag back in the day out of a race committee’s frustration at a J/24 Midwinters in Miami. I once sailed a world championship that had 14 black-flag starts in 10 races, and besides the significant time squandered, I found the results to be unfair, with the regatta ultimately decided by who was best at staying hidden. This is frustrating to competitors, race committees and organizers
—nobody is happy in the end. Maybe we’ve become so accustomed to this punitive approach to managing our starts that we didn’t realize our sport had a starting problem and we couldn’t imagine a better solution.
That was until RaceSense by Vakaros showed us that there is a way forward. Using its GPS-position sensor-based technology, the Vakaros system “sees” all who are over and those who are covered, or not, and it is, as the cliche goes, a game-changer. Developed initially with the M32 catamaran fleet as a beta of sorts and then put to use with other one-design classes on a smaller scale, the recent J/70 Worlds in Spain was Vakaros’ biggest and most ambitious application yet. We’re talking about 95 boats, most of them with top-shelf pros who know how to push the line, get covered, and get away with it. Now we know, with RaceSense at least, there’s no arguing after the fact.
“If you’re over, there’s no one to complain to,” J/70 world champion Doug Newhouse recently told Sailing World. “There are no black flags, no U flags or general recalls. There’s none of that stuff.”
Mark Foster was helping sight the line at the J70 Worlds. He has eyeballed more than his share of big-fleet starts and echoes Newhouse’s experience from the race committee’s perspective. “Being able to automatically identify OCS boats takes that stress off the race committee,” he says. “Of the nine starts (eight races, one abandoned), they were all P-flag starts,” he says. “In one race, there were 23 boats over. They were all identified, and they all restarted within 15 seconds. With the old way, that would have been a general recall.”
“I was amazed,” says Jeremy Wilmot, who won the worlds with Newhouse. “I’ve gone to a J/70 event where we had, I think, six generals before we got the first race off.”
At the heart of the RaceSense system is the Atlas 2 GPS-based sailing instrument. There is one Vakaros unit on each mark and on each competitor’s boats. They are connected wirelessly and individually to a tablet that the race committee uses to control the RaceSense race-management software. The Atlas 2 sensors boast a 25-centimeter accuracy. Considering that at 6 knots, by my math, we go 25 centimeters in less than one-tenth of a second, that’s likely far better than the traditional line spotter’s eye could possibly detect. And likely the start marks are bouncing around more than that each wave anyway.
But specs are specs, and promises are promises. The only way that I would be convinced of RaceSense’s application was to try it myself and understand how accurate enough it is in real-world action.
Foster is the right guy to ask because he has more large events using the technology than anyone else. “I have used it so much now that I have 100 percent faith in the machine,” he tells me. Taken aback with his level of confidence in the Vakaros units and the companion app, I probed further, and he relayed the experience that made him a believer.
“Early on, I called the line using a hybrid of sighting and using RaceSense. Then I called three boats just barely OCS at the VX One North Americans, but RaceSense disagreed, calling them clear.”
To reconcile the discrepancy, he reviewed his video recording of the start, frame by frame. “It turns out, I got it wrong; they were safe by two frames. At 24 frames per second, that’s one-twelfth of a second.”
Foster says that he’s unable to quickly and accurately process the audio of the countdown and combine it with what he sees. He still sights the line—not to call boats, but to make sure that “the number of boats I estimate are over matches how many the machine calls. I have yet to have a recall because of any disagreement from what I see to what the machine calls, but I am prepared to,” he says.
The starting line is defined by GPS units located at each end and, naturally, these units move with the wind and waves. It is the same as a traditional line. This starting information is broadcast in real time to each boat’s Atlas 2 unit. To ensure an even playing field, each boat has the unit mounted in the same place on the boat, usually behind the mast, with a display facing aft.
RaceSense knows the outline of the boat onto which it is mounted, and combined with position data, delivers all the information to know if any part of your boat is over the line, precisely the same way a line spotter would call it. Whereas a line spotter’s sight is limited by distance, close calls and obstructed views, RaceSense knows where every boat is, no matter how well-hidden. And hidden boats are the problem; they are the root cause of all big-fleet starting issues.
At “go,” any OCS boats are instantly notified via a red LED on their Atlas 2; non-OCS boats get a comforting green LED. Returning, the red LED turns green when clear. Meanwhile, the RaceSense software is in the loop and passes all OCS information to the race committee’s tablet.
Classes have a choice in how they want to use the technology. Your class might already embrace pinging, so you will likely already have the equipment and are familiar with using the displayed time, distance, and speed information. For you, RaceSense will be only the incremental addition of automating the process.
But your class might be more grassroots and not want any technology that helps you start your boat. Classes such as the Melges 15 and Etchells have been using it extensively with large fleets. They still want sailors to get line sights and judge their own line, but they also want the technology to improve the start-line experience with better OCS identification. I’m not advocating for either way; it’s a class-management question. Fortunately, your class can inject as much or as little of the technology as desired to match your starting culture.
My first experience using RaceSense at the recent J/24 Worlds in Seattle gave a feel and appreciation for how it works. First, I set up the Atlas 2 with its companion phone app. Because each class gets to choose which features it wants enabled, the app prompted me to input the boat I was racing. J/24 class rules allow full viewing and access to the essential functions: compass, distance to the line, time, and speed information, so it automatically enabled all that. It also uses the J/24 hull outline to know if any part of the boat is over.
I configured the user interface through the app to display this data the way my brain wants to digest it. I chose to keep it simple, with time up top, distance in the middle, and speed at the bottom. I did that all at home. Then, once I got to Seattle, the last step was to register the device with RaceSense for the Worlds, and I was ready to go.
On race day, because the Atlas unit is more than just a starting device, as we sail out to the course, we collect tactical compass headings that are displayed on the Atlas 2. As we get in range of the RC, a “Joined the race” message on the screen automatically alerts us that we have checked in with the race committee, which eliminates coming within hail along with a bunch of other boats trying to do the same.
Later, while continuing our prerace homework, the countdown clock automatically starts, having been triggered by the race committee via RaceSense. For the first race of the day, the clock could be rolling as early as 40 minutes before the start, but between races, it typically started the countdown at 15 minutes. If there was a postponement during the sequence, the clock would stop, then restart when back in action. All this was extremely helpful because it took out all the guesswork and the stress of being near the boat waiting for signals. We always knew how much time we had to make a repair, change our rig, eat lunch, tune up, or whatever we had to do. And we never were off a few seconds, or worse, missed the first signal.
We considered it to be game time when the start clock reached 10 minutes; this is when the race committee hoisted its orange flag, signaling that the line was set. The line information was instantly available on our Atlas 2 unit, with no pinging required.
No one is a fan of ping parade, including Wilmot, who says: “When you’re pinging the line, you’re banging into boats. You’re not looking at the racecourse, and you’re worried about getting back to the end of the line that you want to get to.”
Occasionally at the J/24 worlds, there would be a significant windshift, and the race committee would hoist the AP flag, pausing the sequence to reset the line.
“I was able to move the pin and get back into sequence in eight minutes,” Foster says. “Your line was about 0.35 nautical miles long, and it would have been 30 minutes with repinging.”
Besides being a hassle, pings have suspect accuracy. Sometimes it’s user error: approaching the start mark too fast or from too far away, which results in easily being a meter or so off.
But that error can be dwarfed by natural movements. Jake Keilman, co-founder of Vakaros, was on-site for the J/24 Worlds and tracked that movement. Puget Sound is 600-plus feet deep, and with thousands of feet of anchor line, Keilman reported that the race committee moved 30 meters or more, stretching downwind in the puffs and creeping back in the lulls. He tells me that is normal for deepwater venues, which gave me greater insight into why line sights or pings seem off at times. With pinging, we know only where the line once was, but RaceSense tells us where it is right now. Because that’s how lines are called, that’s quite compelling.
Fortunately, I had no OCS starts at the Worlds, but I had two at the pre-Worlds as I got used to using the system. In each case, my Atlas 2’s red LED gave me instant notification, so I restarted right away. Ten seconds or so later, a green LED told me that I was clear, then I was on my way. It was perhaps just a little worse than a third-row start. It was not a good race, by any means, but it did not take me out of the regatta like the U or black flag would have. And it was a fair penalty—painful, yes—but I still got to sail the race and save points.
Whether I use line sights or a ping, I am never quite sure where the line really is. I often lose my line sight in the last moments, as sails converge and block my view. Even if I can see my line sight, or I have pinged, with the expectation that the line has likely moved since set, I don’t quite trust it all the time. With all that uncertainty, in the final moments of the start, I go with the fleet, desperately trying not to be second row.
“If someone goes, you go with them” is Wilmot’s perspective of the old ways. Hence, boats are often over in clumps, with the inside boats hidden. Even one boat over near an end of the line eclipses the entire line. This leads to either a recall, or if the RC chooses to let it go, free passes for guilty parties. Recalls kill time; free passes are unfair. I’ll say it again: Hidden boats are the root cause of all the other big-fleet starting issues.
Nothing is perfect, and in asking around, the main concern I heard from my friends was cost. It’s all about choice, and your class or event will have to decide if all the benefits are worth the cost. Doug Wake, director of marketing at Vakaros, points out, “If your class already allows pinging, the cost to add RaceSense is relatively small.”
I agree, and for my J/24 Worlds experience, the amount spent on the system was far outweighed by the heightened starting experience. Still, I already owned a perfectly good Velocitek ProStart, but I could not use it because RaceSense is proprietary. I had to buy an Atlas 2. On top of that, there is an additional cost built into the entry fee to use RaceSense race-management software. The business model is evolving, and Vakaros is moving to a subscription-based model, which is a good option for those who use it often enough.
But I don’t think that some dinghy classes are ready to spend. Many don’t already have devices, so they would have to start from scratch. That said, I really look forward to this technology trickling down to the broader base. Some one-design classes likely won’t want all the full features of competitors seeing the line through a machine, but they can surely benefit from automatic and accurate line calling by the race committee. I was glad to hear that Keilman, as a dinghy sailor himself, envisions it going that way. “Our goal has been to try to make it as affordable as we can, and ultimately, as a company, we succeed by getting lots and lots of events and classes to use it.”
I heard a recurring theme of three technology-based concerns: technology creep, cheating, and reliability. Technology creep is fortunately easy to avoid if the adopting class does not allow a few vocal sailors to lead them astray. A class that does not already embrace pinging might be tempted to use all the features once they have the device on hand. I think that it is only a matter of time before less-expensive devices that have no display and show only OCS information become available.
Besides being a hassle, pings have suspect accuracy. Sometimes it’s user error: approaching the start mark too fast or from too far away, which results in easily being a meter or so off.
With regard to cheating, I have heard no signs of tampering, but what if someone played with the placement or hacked it? I would hope that our fellow sailors would not, and safeguards are built in to help prevent it, but still, it’s something to be wary of.
The third concern, with regard to errors, is a valid one, Both the J/70 and J/24 world-championship races went off without a hitch. Yet, we know that salt water is unkind to electronics. Any system component malfunction can be solved by having spares on hand. For bigger issues, such as a software bug or the satellite-provided GPS system going astray, the regatta’s SIs need to be written to enable the race committee to call the line the old-fashioned way.
Our focus has been on starting applications, but RaceSense really is a complete race-management tool to track the fleet around the course, all the way to calling finishes. Its live tracking can be used for publicity, and the telemetry data it provides can be used for training analytics. For example, at the J/70 Worlds, Newhouse and Wilmot agreed to share their individual data through a shared coach who, Wilmot says, “analyzes all our races. I think we had 40-something boats participating.”
I think we will see it used as an aid to both sailors and the jury to make our sport more fair as well. “If you strip it back and look at the underlying cause, it’s a disagreement of when we enter the zone,” says Keilman about most mark-room conflicts, “so our thought is for classes that want to use it either as a learning tool or even instant zone notification while racing, to provide that information and let the sailors sort it out from there.
“And a jury might want it to straighten out the facts.”
I can see tracking information with boat outlines being quite revealing and will help make jury decisions quick, clear, and fair. And likely, upon seeing the facts on a screen onshore, many protests will be settled between competitors over a beer before they even reach the room.
Vakaros is the only automated start system on the market to date, but competition is coming. Announced with a planned mid-2025 launch date is Velocitek’s Wireless Race Management System.
According to Velocitek’s Charles Swanson, Velocitek’s system “delivers synchronized start sequence, automatic line pinging (updated every second), OCS calls with RTK GPS accuracy (1.8 cm), auto-clearing detection, live tracking, and replays.”
Whereas the Atlas 2 is a single unit, Velocitek’s unit will be a split system. Its RTK Puck GPS sensor will be placed on the stern to get it away from the error-causing shielding of the mast, boom and rigging. And you will have a choice of devices and location for your display, from a full screen mounted anywhere you want to a simpler display on your wrist. For small classes with not much room and that want only OCS notification, the wristwatch idea is intriguing.
With a compelling need, line-calling technology is destined to become integral with our sport. Unfortunately, so far they all appear to be proprietary. Your class will have to be on board with one manufacturer or another, and I see no sign of that changing. Worst case, if you race more than one boat, you might need to own more than one if your classes choose different technologies. Just a few months ago, I admit that I was skeptical of the viability and was a bit annoyed at having to pony up the cash, but my experience using it was all positive. I was surprised at how many advantages an automated start-line aid to the race committee brought to the table. And who knows—maybe without needing to blend in to avoid the line spotter’s eye, colorful boats will come back into fashion.
