From Swimming to the Ocean
In my previous article, I shared the story of NEREUS, a cognitive engine that began with a simple question:
Can a machine truly understand movement in water?
For eight years, that question guided my work.
It started with swimming.
It started with SwiMMinD.
It evolved into ZLiX.
But somewhere along the way, I realized something important.
The challenge was never really swimming.
Swimming was simply the first environment where NEREUS could learn.
The ocean was always the bigger challenge.
A Swimmer and a Subsea Cable Have More in Common Than You Think
At first glance, swimming technique analysis and maritime security seem completely unrelated.
Swimming Intelligence
One application focuses on understanding athletes, movements, techniques, and skills in the water.
Maritime Intelligence
The other focuses on autonomous vehicles, underwater infrastructure, environmental uncertainty, and maritime security.
Yet both share the same fundamental problem:
How do you create reliable understanding in an environment where visibility is often poor and information is incomplete?
A swimmer creates turbulence.
Bubbles hide movements.
Reflections distort reality.
The ocean does exactly the same thing, only on a much larger scale.
If NEREUS struggles when bubbles hide a swimmer's leg, imagine the challenge of understanding what is happening around a subsea cable, a pipeline, or an autonomous underwater vehicle operating in murky water.
That was the moment I realized that solving swimming is only a small part of a much larger problem.
Why ORCAPTOR Exists
Europe depends heavily on underwater infrastructure.
Yet the underwater domain remains one of the least understood operational environments.
Communications are limited.
Visibility is poor.
GNSS does not work underwater.
Environmental conditions change constantly.
Even identifying what is happening around critical infrastructure can be extraordinarily difficult.
This is the challenge ORCAPTOR was created to address.
ORCAPTOR
Disruptive Underwater Intelligence and Swarm Technologies for Maritime Security and Critical Infrastructure Protection
The vision is not a single underwater robot.
The vision is an underwater intelligence ecosystem.
From Individual Vehicles to Collective Intelligence
Most underwater systems operate as isolated platforms.
Collect
They collect sensor data during predefined missions.
Execute
They execute instructions established before deployment.
Report
They depend heavily on later human interpretation.
ORCAPTOR investigates a different approach.
What happens if multiple underwater vehicles operate as a collective system?
What happens if they share knowledge rather than simply share data?
What happens if they can adapt together?
ORCAPTOR investigates distributed autonomy, swarm intelligence, resilient underwater communications, predictive digital twins, cooperative navigation, energy-aware operations, and multimodal intelligence.
ORCAPTOR explores how underwater systems might reason collectively rather than operate independently.
NEREUS Becomes the Cognitive Engine
One of the reasons ORCAPTOR is deeply personal to me is that it extends a vision I have already been pursuing through NEREUS.
Within ORCAPTOR, NEREUS becomes the cognitive engine responsible for transforming underwater observations into mission-ready intelligence.
Rather than relying on vision alone, NEREUS combines multiple sources of information.
Visual sensing
Acoustic sensing
Navigation information
Environmental information
These information streams are combined to build an operational understanding of what is actually happening underwater.
The goal is not simply detection.
The goal is understanding.
What was observed?
How reliable is the observation?
Does it matter?
Should the system investigate further?
Should it communicate the information to other autonomous agents?
These are fundamentally intelligence questions rather than sensor questions.
Learning from Nature Again
Interestingly, ORCAPTOR takes me back to the same inspiration that shaped NEREUS.
Nature.
Dolphins do not rely on a single sense.
Whales do not rely on a single sense.
Neither should underwater intelligence systems.
Nature solved many of these problems millions of years ago.
The ocean rewards systems that combine multiple streams of information into a coherent picture of reality.
That is exactly what ORCAPTOR is investigating.
The Bigger Vision
ORCAPTOR is not about building a finished operational product today.
It is about creating the scientific and technological foundations for future underwater intelligence capabilities.
ORCAPTOR brings these capabilities together within a coherent underwater intelligence architecture.
Move from isolated underwater systems toward collective underwater intelligence.
The Journey Continues
When I won the Google, International Olympic Committee and Alibaba Cloud competition in 2018, I believed machines would eventually learn to understand movement in water.
Today, that vision is taking two different forms.
ZLiX
ZLiX applies NEREUS to swimmers and helps people understand, verify, and improve aquatic skills.
ORCAPTOR
ORCAPTOR applies NEREUS to autonomous underwater systems, collective intelligence, and the ocean itself.
One helps people improve their swimming.
The other explores how autonomous systems may one day protect critical underwater infrastructure and support maritime security.
At first glance, they seem like very different projects.
For me, they are the same journey.
Both begin with the same question:
Can a machine truly understand what is happening in water?
And I believe we are only beginning to discover the answer.
Continue Exploring the ORCAPTOR Vision
ORCAPTOR is exploring the foundations of a distributed underwater intelligence ecosystem combining NEREUS-powered perception, collective autonomy, resilient communications, predictive digital twins, cooperative navigation, and energy-aware operations.
Valentin Milkov
Founder of Maze-M and ZLiX
Creator of the NEREUS cognitive-engine concept
Coordinator and lead system architect of ORCAPTOR