Help at the Helm
Autopilots show the way to the future by solving the problems technology creates.
If you take a moment to think about what boats were like even just ten years ago, you’ll realize how much simpler today’s boat systems are to use, pretty much across the board. That simplicity is the result of exponentially deeper complications behind the screens. Now let’s go beyond the electronic end of the equation—when we consider systems with real-world interfaces such as those necessary for autopilots, all bets are off.
In reality, the best autopilot is the one where you don’t feel it at work: No excessive “S” turns as the course is corrected. No matter which way we steer our boats, whether using cable- or hydraulic-ram-driven rudders, stern drives, or the new steer-by-wire pods, the autopilot has to process all the information from all sources—the user, compass heading, the next waypoint on the chart, rudder-feedback sensors, and more—to make sure it stays on course, and deal with course changes quickly and cleanly.
“Boat users expect flawless performance from an autopilot, but without having to interact with it and perform a complicated setup every time they put to sea,” says Ian Matt, Senior Global Product Manager for Raymarine (www.raymarine.com). “The ‘arms race’ between vessel hull design, owner’s performance expectations, and autopilot control action modification has been accelerating to an inevitable point. We have reached that point, where a traditional approach to autopilot control simply will not keep pace any longer.” The challenges for the autopilot designer multiply when you picture the world of boats out there. It would be one thing to design and program an autopilot specifically for just one boat. But that would be a custom system, prohibitively expensive. Instead we have these systems—and every major marine-electronics manufacturer has a skin in the game.
“Every boat is different,” says Dennis Hogan, product manager for Navico USA (www.simrad-yachting.com). “So when an autopilot system is sitting on the shelf in the warehouse, it’s like having a race driver suited up but he doesn’t know what kind of car he’s getting into—it could be a Formula 1 racer, a NASCAR stock car, a rally car, or something else. So he needs to be an expert in all fields.” Think about it, an autopilot could find its way onto the helm of a single-screw displacement trawler, or a stern drive-powered express cruiser, or a pod-drive motoryacht.
Recent improvements in autopilots don’t just hit the latest high-tech vessels out there. The fact that autopilots are getting simpler to operate means that aftermarket sales will eat up a chunk of the production run, and the latest autopilot technology can sometimes wind up on older vessels. The range of equipment in play presents unique challenges. “The equipment on the vessel might not be perfectly up to par, and it might not work as well as it did when it was first made,” says Michael Ellebrecht, senior software engineer for Garmin International (www.garmin.com). “So if you have a leaky helm, and you put your autopilot pump in a hydraulic system like that, it can cause undesirable behavior. Since the autopilot is newly installed, people are going to blame it, whereas the helm might have been the problem the whole time.”
In the world of helm electronics, the autopilot should be top priority, because it actually controls the movements of the boat, rather than just relaying information to the driver as chartplotter, radar, and sounder systems do. “The choice of a pilot is often even more important than the MFD,” Hogan says.
One way that autopilots are integrated into boat systems is through a NMEA 2000 interface. “From an engineering standpoint you look at the CANbus that NMEA 2000 is built upon and it’s a very good reliable communications system,” Ellebrecht says. “It was originally developed for agriculture and automotive applications; it’s fault-tolerant; it’s a very good network. But the problem is in automotive applications and other purpose-built one-manufacturer designs, they know everything that’s going to be on the bus beforehand and nobody adds new functionality. But in the marine world with NMEA 2000, it often ends up being the boatbuilder, who assembles all the different components from different electronics manufacturers.”
The challenges of using such a system for an autopilot are numerous—add in the NMEA 2000 protocol and it could be a volatile combination. “It’s just a question of ‘Has it ever been tested in that configuration?’” Ellebrecht says. “You have all these devices that work well on their own, and then you put them together and it’s not guaranteed that they’ll work well.” Even with the extensive testing a system such as an autopilot goes through before being released to the boating public, one cannot know all the factors existing on all the boats where the system could be installed.
“To handle the larger steering loads that we’ve been seeing more of with the larger engines, we came out with our GHP20 with a SmartPump,” Ellebrecht says. “The SmartPump has a much more powerful motor in it, and a bigger electronic amplifier. It can handle the larger steering loads without overheating. So it just provides a more reliable hydraulic steering solution over a broader range of vessels.”
As boaters consider upgrading their electronics, they need to consider the implications of their autopilot needs. “Today, more folks are linking MFDs to autopilots to get a fully integrated system,” Hogan says. “Simrad has recognized this and encouraged even more tightly integrating the MFD with the pilot.” Simrad does not require a separate autopilot control head, instead operating the system as a window within the multifunction display. According to Hogan, the result is a system that’s easier to use and makes better use of valuable helm dashboard real estate.
The newest autopilots are designed to work with the latest onboard systems—steering, propulsion, and more—and use the boats’ own systems to make turns and course adjustments. But these systems face their own challenges. “With the drive-by-wire steering systems, response is both immediate and fast,” Matt says. “From a control point of view, that presents a whole new challenge. Now an autopilot no longer has to push hard to effect a rudder movement, it actually has to act very gently to avoid the steering swinging right past the desired angle, and inducing the dreaded ‘essing’ steering instability. This would be OK if all the drive-by-wire steering manufacturers implemented their systems in the same way. But each has its own unique interface, unique handling properties, and its own unique problems which often fall to the autopilot to correct for.”
Still other developments show the direction things are headed. Furuno’s NavPilot 700 (www.furunousa.com) uses a software program called Fantum Feedback that allows the autopilot to work with outboard systems without requiring the attachment of an actual rudder-feedback unit.
And then there’s Raymarine. “With the Evolution autopilot, Raymarine’s flipped the control problem on its head,” Matt says. “Within the autopilot, is a model of the ‘perfect boat.’ When a steering change is requested, the Evolution knows exactly how the ‘perfect boat’ would respond. Using its unique and highly advanced 9-axis sensor, the Evolution now measures the actual boat response compared to the ‘perfect’ response and modifies its control action accordingly. In essence, the pilot is now adapting to the environment, rather than to the boat itself. By taking this approach, Raymarine has designed an autopilot which will ‘evolve’ its steering action every second, of every minute, of every hour it is in use.”
As autopilots learn how to solve the problems we present to them, users need to learn to let them do their thing. And the next thing you know, your boat will be taking you exactly where you want to go.
This article originally appeared in the November 2013 issue of Power & Motoryacht magazine.