SubOptic 2023 Shark Tank! A Guest Blog with Steve Grubb of Meta

Coming up with original and entertaining ideas for sessions within an established event is tough. But our guest for today’s blog, Meta’s Global Optical Architect, Dr. Steve Grubb, really scored a home run for one of his sessions at the recent SubOptic Conference in Bangkok, Thailand.

Steve came up with the idea of basing a session on Shark Tank, a TV show format that’s shown in many countries around the world, sometimes called Dragon’s Den (at least here in the U.K.). Whatever it’s called, the show format has a number of entrepreneurs pitching their ideas to a panel of Sharks (or Dragons), who quiz the presenter on the practicality of the idea and, if they like it, offer funding and support.

I was honored to be asked to be one of the Sharks on the panel, along with Valey Kamalov of the SubOptic Foundation, Olivier Courtois of Alcatel Submarine Networks, and Elaine Stafford of DRG Undersea Consulting.

We heard a total of six pitches from Kitch Kennedy of Saildrone, Philippe Perrier of Meta, Sergejs Makovejs of Corning, Pierre Mertz and Antti Kankkunen of Infinera, Takimi Hasagawa of Sumitomo, and Nate Lindsey of FiberSense.

Geoff Bennett: Steve – how did you come up with the idea for the Shark Tank?

Steve Grubb: There are many innovative ideas in submarine technology being worked on that are early stage and not yet suitable for a full paper submission to SubOptic. Hence the conference would have no visibility at all into this work. This event provides a perfect forum for these great ideas. We actually had the main plenary hall for the event, and it was pretty much a full house. So if we’d had an interactive Q&A or round table approach, with so many people and so many innovations, we’d probably still be there now! As it was, I think the presenters all did a great job with their pitches, you Sharks raised some great questions, and we could stay on time, even with an interactive format. And let’s face it – who doesn’t love sharks?

GB: So true – I think they’re delicious with a Cajun-style spice crust! So, maybe you could give us a quick summary of the six pitches.

Pitch #1: Submarine Cable Survey by Drones

SG: OK, so Saildrone was first. One of the most time-consuming parts of a new cable deployment is a route survey. The traditional approach uses a ship full of crew, burning diesel and throwing a lot of CO2 into the air, and is quite expensive. It’s a costly process in more ways than one. Saildrone already has a fleet of remote-controlled survey drones. They’re powered by solar-charged batteries and a sail too, so they’re basically non-polluting. You get a ton of useful data that can be viewed in near real time – especially as the next generation will introduce Starlink satellite connections.

GB: It sounds like a great idea – but they’re already a successful technology. Do they need us Sharks?

SG: They’ve only been funded for ocean mapping and scientific applications. Broadening their market to subsea surveys requires additional funding and resources to develop this application-specific autonomous vehicle. I think they deserve to succeed and it’s a really cool idea.

Pitch #2: Mid-ocean Power Source for Submarine Cables

GB: OK, moving onto the Meta pitch with Philippe Perrier. What’s his idea?

SG: It’s a mid-ocean platform that takes solar and wave power and delivers it to a subsea cable to offer a mid-span power boost. Basically, it could give us a new dimension to scale cable capacity by removing power constraints. It could also be an emergency power source in case of cable shunt faults.

GB: And I seem to remember that it needs a liquid ballast anyway, and as a Shark I suggested that filling it with beer and using it as a sport fishing platform could give us a new revenue stream – right?

SG: I just assumed you were jet lagged when you made that suggestion. You mean you were serious?

GB: I always like to keep people guessing if I’m serious or not. OK, so what do you think of the mid-span power idea?

SG: You know, it really does address a critical need in long-distance subsea cables. There are some issues we’d need to iron out – like how will the cable react to the movement of the platform in rough seas? But a company called GEPS Techno is pressing the project ahead. They’ve deployed platforms for testing for the oil and gas industry and they’re looking to deploy on real submarine cables by the end of the decade.

Pitch #3: Multi-core EDFA with Core Pumping

GB: Moving onto our next victim…err…entrepreneur. That was Sergejs Makovejs from Corning. And I seem to remember he was one of two people pitching a multi-core EDFA for a future generation of SDM cables.

SG: Yes. It’s pretty certain that we’ll eventually see multi-core fibers appearing in SDM. Multi-core fiber addresses the issue of spatial scaling – fitting more fiber cores into the existing 17-mm cable size that’s been tested and certified for submarine deployments. Increasing that cable diameter is nontrivial for various reasons. But multiple cores would normally imply a fan out and fan in at each repeater and a separate EDFA amplifier for each core. Sergejs’ pitch was to develop a single EDFA that pumps multiple cores at the same time. That saves on equipment costs and also has better electrical efficiency than using separate EDFAs.

GB: But what if there is crosstalk between the cores? So we’d have to implement a compensation based on multiple input multiple output (MIMO) in the coherent receiver?

SG: Yes, that’s quite likely. To be honest, a MIMO receiver is going to become necessary if multi-core fibers scale to four cores and beyond anyway, because higher-core-count fibers are inevitably going to introduce signal coupling between the cores. Right now that’s a long way into the future and we have other ways to increase cable capacity before we need to consider MIMO in the transponder. Let’s see how that one pans out.

Pitch #4: Long-range Seismic Detection on Existing Submarine Cables

GB: OK, next up we had my colleagues from Infinera, Pierre Mertz and Antti Kankkunen. I had to recuse myself from this one because of a conflict of interest!

SG: And rightly so! But it’s a great idea – to turn existing submarine cables into subsea earthquake detectors. Pierre has done trials on a couple of cable systems, and the results are promising. But it needs further development and testing – and a commitment to put the seismic detection functionality into a coherent ASIC. And Antti’s additional pitch was to see if the same technique could be used to detect other vibrational events that affect the cable – like ships’ anchors or fishing nets dragging close by. Together those events cause most of the cable outages in submarine cables, so if we could detect them and try to warn off those vessels it could be very useful.

Pitch #5: Multi-core EDFA with Cladding Pumping

GB: And next up we had another multi-core EDFA idea, this time from Takimi Hasagawa from Sumitomo. How did it compare to the Corning idea?

SG: Corning’s multi-core EDFA was based on a core-pumped idea and the Sumitomo approach uses cladding pumping. There are pros and cons with each approach, and I think it’s interesting to see how those pros and cons differ between single-core and multi-core fibers. But, just like the Corning EDFA, Sumitomo is talking about coupled cores – which means that the transponders would have to implement MIMOs.

GB: And is MIMO a difficult technology to implement?

SG: Difficult? Well, not really…no more so than anything you do today in the latest optical engines. But the key question is whether there’s enough of a market in subsea alone to justify MIMO development. The key here is that there’s no obvious short- or medium-term use case for multi-core fiber in terrestrial networks, although that could change in time as the scaling wall hits us. There is tremendous MIMO development for 6G mobile technology, which has a very similar development timescale to multi-core cables, although this is at much lower frequencies than optical transmission. Perhaps the algorithms and libraries developed for the 6G market could be leveraged. But if transponder vendors have to develop MIMO purely for subsea applications, then it’s unlikely that this will be economical – at least that’s the way it looks today. In five or 10 years’ time it might look very different if traffic continues to grow the way it is.

Pitch #6: Short-range Seismic Detection on Existing Submarine Cables

GB: And our final pitch was by Nate Lindsey of FiberSense. So that one is another earthquake detection technique – this time using distributed acoustic sensing, or DAS. How does that compare with other techniques?

SG: DAS uses a kind of coherent optical time-domain reflectometer (OTDR) approach – it sends a powerful and very narrow laser pulse into the fiber, and it measures the back scatter from tiny imperfections along the fiber path. DAS is extremely sensitive – you can pick up all sorts of phenomena, including passing whales, surface wave motion – pretty much anything that makes a noise.

GB: Is there a but here?

SG: To some extent, yes. DAS has a limited range today of about 50 km from each end of the cable, the first span. It might be possible to extend that range to 100 or even 150 km, but that’s hard to do and adds cost to the equipment. Even so, DAS offers excellent detection over that first part of the cable, and these are the sections that are often affected by anchors or fishing nets. So I would say it could be a complementary technique to Infinera’s suggestion – which covers the full cable length, but at much lower resolution.

GB: Steve – I thoroughly enjoyed the Shark Tank at SubOptic. I have a feeling you’ll be asked to do this again at the next event in Lisbon in 2025. And I hope you’ll consider me as a Shark once again.

SG: Well, you’ll need to be on your best behavior next time! I knew you would be a wild card, and you did not disappoint.

GB: It’s a deal. Thank you for appearing on our guest blog!

Dr. Grubb is currently a Global Optical Architect at Facebook, overseeing the build of several new open submarine cable systems and introducing new optical technologies for Facebooks global network. Prior to Facebook, he was a Fellow at Infinera where for 14 years he directed work on next generation Photonic Integrated optical and network technologies. He was also responsible for the first commercial introduction of Raman amplifiers in fiber networks. He received his Ph.D. from Cornell University. Dr. Grubb has over 100 published papers and conference contributions and over 75 issued US Patents.