Simrad's AI Sonar, Tested: A North Carolina Captain on What the Virtual Operator Catches
Big shout out to Huk, our title sponsor, for powering this episode of The Science of Fishing. Huk is the show's title-sponsor partner; that arrangement is what keeps these conversations free for our readers and listeners. Sponsorship does not shape our editorial — guest selection, topic framing, and product coverage are independent of any sponsor relationship.
Simrad's new AI is a second pair of eyes on the sonar screen — and Captain Greg Mayer keeps seeing it mark fish he was about to look for himself. On a slow charter day running back to Oregon Inlet, North Carolina, Greg and Simrad's Martin walked The Science of Fishing through what a "virtual sonar operator" actually does — and where it's headed next.
What a virtual sonar operator actually does
Greg's first reaction was wariness. He expected an autopilot gimmick — something that would fly the spread for him, or worse, push the boat onto a number it thought was the right one. "I'm like, okay, AI is going to take over my phone or it's going to bring the boat, put it right on the perfect spot to get a bite. I said, no, that's not going to happen," he told us at [01:46]. "But now that I've used it a little bit, it's actually a pretty cool tool."
Martin's framing for the listener in the truck is simpler than the marketing copy. "You actually have an assistant with you. He's just virtual. He lives inside the machine, and he does what any sonar operator basically would do. He's just staring at that screen constantly, looking for the fish," he said at [02:23].
Concretely, the AI drops a marker on the screen — but a visually distinct one, so Greg can tell at a glance which marks are his and which the AI's. "When Captain Greg drops a marker, it's outside down, it's a triangle, and then we invert the triangle when the virtual sonar operator does it," Martin explained at [03:43].
There's also an audio channel that most of the Western sportfishing fleet has forgotten about — the raw acoustic return from the ping. "There's a lot of information in the audio of the [ping] and the return of the fish," Martin said at [04:16]. "We can't listen to that for 12 hours out fishing, but the sonar — the virtual operator — can actually turn on the audio just when he sees something. So if you connect the speaker, you can actually … hear an alarm too that is actually the echo of the fish."
How the AI learns: the captain marks, Norway annotates
Asked how the system separates a fish from clutter, bait, grass, or thermocline, Martin pointed back at the human in the chair. "We observe what they do. So when Greg puts down a marker, the AI kind of understands, okay, so Captain Greg [thought] this was interesting. This is worth something. And that's [a] way of training it," he said at [05:40].
The verification loop sits in Norway. "We get this data back, and myself and a few of my guys in Norway and also around the world, we go into the data and we verify it and annotate it. So that's how we [learn] continuously and get better, better every day," Martin said at [06:03].
Greg's spot stays Greg's spot — more on that below — but the model itself benefits from every captain who runs it. And the road ahead isn't just "better pattern match." Martin pointed out that the virtual operator has something the human at the wheel does not: time. "The AI also has a lot of time on his hands and a lot of processing power. So going forward, the AI also has a potential to go deeper into the data," he said at [06:34].
Greg in real conditions: from "marker confirms" to fish he'd have missed
Day-to-day, Greg told us he uses the virtual operator two ways.
The first is confirmation. "A lot of the time I'm watching the screen anyway. So if I see something, I'll move a marker over there and there've been several times the AI [marks at] the same time I am. So it kind of confirms to the AI that there is a target there," he said at [07:06].
The second is backstop — for the long stretches he's not on the screen at all. "I'm fishing four rods up here, two teasers. I'm topping on my customers and there's a lot of things going on. So I'm not always looking at the screen," he said at [07:32]. "Then I'll turn around and I can hear the alarm, and I could see the target on the screen. And then it gives me an idea of where I need to start looking."
The "I would have missed that fish" moment, when we asked for one, came quickly. "Several," Greg said. "We were doing a podcast a couple of weeks ago, Martin and I … I got a bite and Martin, when he was watching my screen — I wasn't watching — and it just did put a marker on there. … I sort of moved over to it and there [it] was." [08:44]
For the amateur or weekend captain, Greg thinks that backstop is the real prize. "Now you're getting an AI that's getting years of experience much quicker than any of us could get it because they're pulling data from … all the different units," he said at [15:02]. "The average boat owner is probably fishing maybe 15 days a year. We're out here 150, 200 days a year."
Tuna vs. marlin: where AI sonar earns its keep
The reflex assumption is that sonar belongs to the billfish crowd. Greg pushed back on that. "For me, it's more effective for [tuna]. It's a much bigger target for the sonar to find," he said at [10:25].
And once the tuna show up on the sonar, the second-order effect is that everyone's on them. So Greg uses the AI to fish away from the fleet. "The way I use it a lot is … I go out on my own and look around. Once you get in the fleet, when everybody [is on the] school of fish, they're constantly getting pursued. And if you're off on your own a little bit, you get a couple more chances to get some bites before they get spooked," he said at [10:59].
Martin noted the tuna market wasn't the original target for the hardware. "This market was a little new for us back in 2021 when this sonar came out — the tuna market with the purse seiners and so on," he said at [11:38]. The AI work is built on top of that hardware base.
For more context on how omnidirectional sonar has reshaped the offshore game generally, InTheBite covered the omni-sonar shift in 2024.
Tournament fairness — and the "if you can't pay, don't play" debate
Mark pushed the obvious next question: is AI-assisted target detection just the next link in the electronics arms race, or is it something tournament rules will eventually have to address?
Greg's answer leaned on history. "When omnidirectional sonar first got introduced into the fishery, every corner of it changed because of it. And I know several captains that were vehemently opposed to it … And now they've got a sonar. Now it's all perfectly okay. It's just the evolution of the fishery," he said at [16:26].
Then came the vintage soundbite — Greg recalling a charter captain buddy from 25 years ago who had a $20,000 electronics rig back when that was unheard of, and the punchline he heard at the time:
"If you can't pay, don't play."
— Captain Greg Mayer,
[17:12]
"And I think that's a lot of what's going on with omnidirectional sonar now," Greg said. His prediction for AI sonar in tournament fishing is matter-of-fact. "There may be a boat that starts winning tournaments with it, and somebody's going to have sour grapes to complain. We'll see where it goes from there." [17:35]
Data ownership: your spot stays your spot
One of the surest ways to lose a charter captain is to ask them to upload their numbers. Mark put the question to Martin plainly: who else can see this data?
"It's just for the boat," Martin said at [18:45]. "I've done some Instagrams and YouTubes where I show how we train it and I use, amongst others, Greg's data, but I have cleared with Greg that I can use those data because you could also see the position. … No, your fishing spot is still very much secure. We treat that very confidentially."
There's an opt-in sharing model on the roadmap — but only when both captains say yes. "If Captain Greg and another captain wants to share, we might enable that, that they can share. But [otherwise] you're fishing spot is still very much secure," Martin added at [19:08].
Bycatch, fuel, and the conservation case
Mark steered the conversation toward the angle he cares about most — conservation. Less fuel burned, less time wasted on dead water, and less pressure on the resource.
Martin laid out the long arc that Simrad — operating as Simrad Fisheries inside the Kongsberg group — has been pushing for decades. "We have this slogan with Simrad [Fisheries], we're part of the Kongsberg group. … We've been in the commercial fishery since the 50s with radios and the echo sounders and later sonars. … Once a virtual sonar operator can see across all of [the sonars on board] and link it to your actual catch, then you can really start handling bycatch perhaps and making it more efficient and burning less fuel." [20:43]
Greg's example is one every offshore captain knows: shark depredation. "The bigger the mark you see, the higher the odds that it's more sharks than tuna," he said at [22:01]. "And with the sonar, you can turn away from it … I would rather go around and look for a smaller mark — it's more tuna and less sharks — than go to the bigger ones where I know I'm going to get decimated."
What's next: AI that tunes the sonar itself (the SY50 vision)
Greg's wish list is the Simrad SY50 — Simrad's compact omnidirectional sonar — with an AI that doesn't just mark fish, but also runs the settings panel. "There's several filters, there's several settings — if you TVG your, your RCG, the different [gains] on there — there's so many different settings that you can use," he said at [27:02]. "I'm really looking forward to when AI will be able to look at the water conditions … and from there it would give you your recommendation about what your ultimate settings [are]."
Martin said that's already in build with Viam, Simrad's AI development partner. "The AI is already able to start zooming in, changing using the tilt to actively search. It's not released yet." [12:30]
The end state Martin described is the one Greg actually wants: a captain on the rail, not in front of the screen. "In the future it can look at the data behind it too so it doesn't need the settings, so Greg can get back to watching the ocean instead of staring at a screen." [28:06]
Listen, and go fishing
The full conversation runs about 29 minutes. Stream it on the Science of Fishing podcast page, or watch the video version above. And if you want to see this hardware run on the deck where it's actually being tested, Greg said it best in his sign-off: go fish with him up in North Carolina. He runs the Fishin' Frenzy out of the Oregon Inlet Fishing Center.
Adapted from a Science of Fishing podcast conversation with Captain Greg Mayer (Fishin' Frenzy, Oregon Inlet Fishing Center) and Martin (Simrad / Kongsberg Discovery). Sonar AI product naming and the Viam partnership verified against the Kongsberg / Viam press release (March 24, 2026). Related coverage: InTheBite's 2024 piece on the omni-sonar revolution.
