Marine Electronics - Are You Smarter Than Them?By Lenny Rudow
Published: June/July 2012
It's not HAL 9000 against David Bowman, but the results are just as epic.
"Excuse me," said a bossy and annoying cyber-voice, dripping with the innate overconfidence of a TomTom. "This is your electronics talking. Did you know I can do that now? And talking isn't all. I can navigate your boat all by myself. I can choose my own settings and sensitivities, see miles farther into the distance than you can. I even find fish faster than you. Sorry meat-bag, your days as captain are numbered. I'm now smarter than you, and don't need you anymore."
I sit bolt upright in my berth, heart pounding. Wow, what a crazy dream — my own marine electronics were trying to pull a "Terminator" on me. On the other hand, maybe the idea of a seagoing Singularity isn't so crazy after all. From some of the conversations I've had with industry insiders, you'd think that we humans are no longer the most intelligent entities on a boat. Take the words of Greg Konig, a vice president and product manager for Navico, for example: "The hardware and software we designed for the Broadband radar is so good, so advanced, that it can outthink even experienced boaters. Take the unit off the automatic settings, and see for yourself. I’ll bet you can't tune it better than it can tune itself, in most normal situations."
How dare he insinuate that some dumb dome was smarter than I am! On my boat I'm still the captain, the master, the decider-in-chief, and most certainly the smartest entity (unless my wife is aboard). The gauntlet had been thrown down. I decided right then and there to take a Broadband 3G unit out on the water, and prove in several different situations that human nautical know-how could outperform the latest algorithms and programming. And when Konig threw in a snide remark about a modern fishfinder's digital-processing capacity, I knew I’d have to test animated abilities against these as well. I'd prove who’s better, the man or the machine. And in doing so, discover a thing or two about what this kind of modern technology means to us, as boaters.
Break Bad On Broadband
I've used plenty of modern radar in the past, and have yet to find one that could even come close to outsmarting me. So, what's different about Broadband? Several features change the way automatic adjustments are made. For starters, unlike traditional pulse radar, Broadband uses a different STC (sensitivity time control) curve (for adjusting amplifier gain) for every range, specifically optimized for that particular range. Secondly, because the unit is solid-state, there's no tuning necessary on startup. Third, signal processing takes place at a higher speed than traditional pulse radar, with 16 bytes of data as compared to 10 or so. On top of that it has more data to work with since it's operating at 75 MHz of bandwidth, as compared to 18 to 20 MHz with more common radar. There's more — but Navico techie Don Korte would only give away so much information before shutting the conversation down.
To put this technology to the test, I started on short range (1/8-mile), in an area with plenty of hard targets: a tiny creek off Selby Bay, surrounded by tall trees and large homes. With the radar on standby, I put the machine on manual and scrambled the sensitivity, rain and sea clutter, interference rejection, and target boost. Then, I started transmitting and got busy with the buttons. A few minutes later, I was pretty darn satisfied with myself. I had hard returns all around, no clutter on the screen, and I could even pick up a pair of crab trap floats bobbing close together next to the boat.
I took a screen shot for evidence, then to prove my superiority, switched into auto mode. Unfortunately, there wasn't a whole heck of a lot of difference in what I saw. When I placed the screen shots side by side and examined them more closely, however, I didn't like the results. The floats (indicated in the screen shot with the blue arrow), which were more or less merged and could be mistaken for a single return on my manually adjusted screen, were much more clearly differentiated by the Broadband's brain. And a feeder creek just a few feet across was actually visible, while on my screen it wasn't. Ouch! Score one for the machines.
The radar does better on its own in long-range views, too; Thomas Point Light (indicated by the white arrow) is much more clearly defined, and two day markers (marked by green arrows) are visible on-screen, instead of one.
Hopefully, I'd have better luck with longer-range views. I motored out into the mouth of the river, re-scrambled the settings in manual mode to get a fresh start, and switched to 12-mile range. With a few minutes of careful adjusting, I was happy with what was on-screen: I could see the small, red day marker a half-mile off to starboard (green arrows on the screen shots), got a nice hard return from freighters four or five miles off dead ahead and slightly to port (black arrows), and I could see Thomas Point Light about two miles out, slightly to port (white arrows).
"Damn, I'm good," I thought, as I switched back into full-auto mode. Seconds later, I dropped the "I'm good" part of that expression as I stared at the screen in disbelief. While manually adjusting, I must have desensitized the radar too much, because in auto mode I saw a second return beyond the day marker, and Thomas Point Light provided a much better target. At least the freighters still looked OK. Just to be sure the current screen was accurate, I used my binoculars to verify the existence of the second return beyond the day marker. Yup, something was there, all right — the green marker #1A, exactly where my chart said it should be. I must've tuned out in manual mode. Bummer! Machines 2, human being 0.
Being a die-hard angler whose experience dates back to the days of flashers and paper graphs, I had high hopes I could do better pitted against the fishfinder. But after speaking with a few industry sharpies on the matter, I knew it would be no easy feat.
"When automatic functions first appeared, a knowledgeable captain could always set the machine up better himself," said Allen Scheider, a vice presidnet at Si-Tex. "But with the advent of digital sounding technologies and high-speed processing capabilities, today the auto functions are just about foolproof." Jim McGowan of Raymarine also believes that going digital makes a huge difference. "Digital signal processing techniques (DSP) are far more sophisticated than the analog products of just a few years ago," he explained. "There are no fixed filters in the receiver (as older analog units had). Instead, our HD digital fishfinders send all the information to the receiver and let the DSP chipset (a high-performance computer chip with custom software) sort it all out. This makes the filtering totally dynamic."
Even if you think you can do better than the fishfinder's automatic settings, there's a downside to using manual mode that one must consider: In many cases, it means you'll have to constantly shift the range scale as depth changes. Some new units from Garmin get around this issue and split the difference between manual and auto, with "hybrid" auto-settings (auto-high, auto-medium, and auto-low). Furuno also has a way of addressing this issue. "We've found that as good as the auto features are, seasoned fishermen still like to be able to manually tune in certain situations," advertising and communications manager Jeff Kauzlaric said. "So we separated our auto features. This way, an advanced bottom fisherman can use manual gain while taking advantage of auto range control."
Hurricane Irene Creates A Test Bed
Truth be told, I know from using the latest fishfinders from these manufacturers and others that, generally speaking, on modern high-quality units, the auto mode works extremely well — most of the time I leave it on myself. But unusual conditions can sometimes fool a mechanized brain. And when push comes to shove, I'm pretty confident I can beat the average angling automaton. But there's only one way to say for sure if the human brain can outthink this new technology: It's time for another on-the-water test. After Hurricane Irene passed, the Chesapeake's waters were roiled and churned, full of suspended solids and multiple temperature barriers. That would be a good challenge.
I entered the bay and slowed down to eyeball an underwater hump, and it became obvious that my fishfinder had a very tough time dealing with the conditions. The screen was constantly cluttered, and although I could make out fish hovering over the structure, the screen was quite cluttered and could easily be deceiving. So I switched to manual, and set the meter where my human brain said it was best. In a nutshell, I kicked Cyber-Charlie's butt! I could get the screen 80 percent cleaner without losing the ability to see those fish, and it only took me a minute or two to do so.
Just to prove I knew exactly what I was looking at, I dropped a jig over the side and caught a few stripers. Finally, I could claim victory! That stupid machine doesn't have the brain power to make the right adjustments as quickly and effectively as my good old medulla oblongata.
To Home Page
Lessons LearnedSo, what does pitting man versus the machine tell us? First off, it's pretty clear that Broadband radar does, in fact, have a pretty darn impressive brain. If you have one on your boat, chances are you'll be best served by leaving it on auto mode. What about other brands? I haven't personally experienced any other unit that can self-adjust this well, but it's easy enough for you to test your own radar for yourself. All you need is a clear day, a strong pair of binoculars, and a range of strong and weak targets to choose from. You might be surprised at the results — and even if you're not, chances are good you'll learn a thing or two about how to best adjust your own unit.
When it comes to fishfinders, in average conditions on average days, you're likely to see beneath the surface best in auto mode. But the moment you aren't satisfied with the view, start pressing buttons. Many different anomalies in weather and water composition can deceive these devices, as we saw with the hurricane-riled waters of the Chesapeake. I've also observed modern units being fooled by algae blooms, heavy jellyfish populations, dense schools of fish, and aerated surface waters.
To effectively manually adjust your fishfinder's sensitivity, start by increasing range to double the water's depth. Then increase gain until a second bottom echo appears on-screen. Shift back to the proper range setting, and you have a good starting point. If most of the clutter you see is at the top of your screen, increase the STC setting next. If not (or after boosting STC), reduce gain until mid- and deep-water clutter is reduced. As long as you continue seeing fish and receiving a solid bottom reading, slowly bump the gain down a notch at a time until fish are still present and clutter is not. If you're in deep ocean waters, you lose bottom and/or fish readings as you make these adjustments, and you have a multi-frequency unit, try shifting to a lower frequency and repeating the above process. And remember, if some arrogant, overbearing automated voice tells you that the human brain is outdated, don't believe it. Yet.