How a Data Science Internship Helped Advance AI for Persistent Ocean Monitoring

An inside look at how real-world AI development, hands-on field testing, and cross-disciplinary collaboration helped launch a career in Canada's growing blue economy.

Artificial intelligence is transforming the way we understand and protect our oceans. At Open Ocean Robotics, that transformation happens through autonomous, zero-emission uncrewed surface vehicles (USVs) equipped with intelligent software that helps operators detect, track, and respond to activity on the water in real time.

Behind every advancement in our AI technology is a team of engineers, software developers, and data scientists working together to solve real-world challenges. For Carlos Serrouya, what began as a Data Science Internship quickly became an opportunity to contribute directly to one of those technologies: Enhanced Horizon™, Open Ocean Robotics' AI-powered visual detection system.

Having completed a 12-month term, Carlos reflects on his internship, the lessons he learned, and why building AI for the ocean has shaped his career in ways he never expected.

What is Enhanced Horizon?

Enhanced Horizon is Open Ocean Robotics' AI-powered visual detection system, designed to help operators see more, react faster, and monitor the maritime environment with greater confidence. 

Using 360° optical and thermal cameras, Enhanced Horizon detects, classifies, and tracks vessels, marine debris, and other surface objects in real time, displaying them directly within Open Ocean Robotics’ cloud-based control portal, XplorerView™. By automatically identifying and tracking objects, the system reduces operator workload while providing actionable intelligence for maritime security, environmental monitoring, and autonomous operations.

Enhanced Horizon™ detecting and tracking a distant object. The red triangle on the XplorerView™ mission portal indicates the AI-generated detection.

Finding the Right Opportunity

When Carlos first came across Open Ocean Robotics, he was searching for a data science internship that would allow him to apply machine learning to meaningful, real-world problems.

After discovering the company online, he began researching the team behind it.

"Everyone had incredibly impressive backgrounds," Carlos recalls. "I looked into the company, saw the experience across the engineering and leadership teams, and thought it would be an incredible place to learn."

Originally from Revelstoke and studying in Calgary, Carlos was also excited by the opportunity to relocate to Victoria and work alongside a multidisciplinary team developing autonomous marine technology.

What he didn't yet realize was just how hands-on the experience would become.

Building AI That Works on the Water

Carlos joined the software team with previous experience developing machine learning pipelines but quickly found himself tackling challenges that were entirely new.

Initially, Carlos worked on improving the infrastructure that supports machine learning development and testing, helping streamline the process of moving new AI models from development into production.

As his experience grew, so did his responsibilities.

"I started building algorithms for object tracking and localization," he explains. "The evolution was going from working on software running on local machines to developing code that runs directly on the edge devices onboard our USVs."

That transition introduced Carlos to one of the biggest differences between academic AI projects and commercial autonomous systems.

Building AI models is only part of the challenge.

Deploying those models onboard autonomous vessels operating in dynamic marine environments requires software that is reliable, efficient, and capable of handling real-world conditions.

"I'd never coded on an edge device before joining Open Ocean Robotics," he says. "Learning the software principles, error handling, and robustness required for software running onboard autonomous systems was completely new for me."

He also helped improve how Enhanced Horizon processes detections after the AI model has identified potential objects.

Rather than simply displaying raw detections, the system applies additional statistical processing and tracking algorithms to estimate object locations and reduce false positives, helping provide operators with more reliable and actionable situational awareness.

Rapid Innovation Through Real-World Testing

One aspect of the internship stood out above everything else: the ability to rapidly test new ideas on the water.

Instead of waiting weeks or months to validate new software, Carlos was often able to help launch one of the engineering USVs and evaluate new algorithms within days.

"It's super rewarding to see what you've done last week go on the water this week," he says. "You immediately see what works, what needs to change, and how operators interact with the technology."

As his work on object localization progressed, Carlos began organizing field tests himself, coordinating with operators and collecting data that could immediately be used to refine the algorithms.

"If there was something I wanted to test, I could go test it," he explains. "That level of agility is really exciting because you're directly involved in improving the technology."

Being located just steps from Victoria's waterfront also played an important role.

The close proximity between the engineering office and the ocean allows software teams to quickly validate new features, collect additional datasets, and continuously improve system performance through frequent on-water testing.

For Carlos, that rapid development cycle became one of the defining characteristics of working at Open Ocean Robotics.

A DataXplorer™ USV is launched into Victoria Harbour, where routine testing and mission deployments begin just outside Open Ocean Robotics' headquarters.

Learning Beyond Data Science

Although Carlos joined the company as a Data Science Intern, the experience extended far beyond machine learning.

Each day began with software team check-ins before individual development work, while regular testing sessions brought together software developers, robotics engineers, field operators, and cloud specialists.

"Whenever I ran into a problem, there was always someone who could help," he says. "Whether it was data science, software, cloud infrastructure, or embedded systems, everyone was incredibly knowledgeable and willing to share what they knew."

Working alongside experts across multiple disciplines accelerated his learning and provided exposure to areas he hadn't previously explored, including embedded systems, production software, and autonomous robotics.

Rather than working in isolation, Carlos experienced firsthand how successful AI products are built through close collaboration between software, engineering, and operations teams.

Why Real-World Testing for AI Matters

Developing AI for autonomous maritime operations presents unique challenges that go far beyond building accurate machine learning models.

Software must operate reliably onboard edge computing devices, process information efficiently, and continue performing in constantly changing environmental conditions.

According to Carlos, Open Ocean Robotics' ability to rapidly iterate while maintaining robust software is one of the company's greatest strengths.

"The robustness and rapid iteration we have is only possible because of the incredible people here and because we're so close to the ocean," he explains. "We can test things constantly, gather real operational data, and continuously improve the software."

That combination of experienced engineers, frequent field testing, and real mission deployments helps ensure technologies like Enhanced Horizon continue evolving based on operational experience rather than laboratory simulations alone.

Discovering a Career in Ocean Technology

Before joining Open Ocean Robotics, Carlos already knew he wanted a career in data science.

What surprised him was discovering how naturally those skills could be applied to the ocean.

"I'd always been interested in data science professionally and the ocean personally," he says. "I'd just never thought about combining the two."

Throughout his internship, he found himself becoming increasingly interested in maritime autonomy, embedded systems, and AI applications for ocean monitoring.

The experience ultimately reinforced his desire to continue building technologies that have meaningful real-world impact.

Advice for Future Interns

Looking back, Carlos believes curiosity is one of the most valuable qualities an intern can bring.

Ocean technology combines software, robotics, communications, navigation, environmental science, and countless specialized systems. That means new team members are constantly exposed to unfamiliar terminology and technologies.

His advice?

"Ask lots of questions," he says. "If you keep hearing an acronym or a system mentioned, take the time to learn what it does. Even if it doesn't seem relevant today, chances are you'll end up working with it later."

That willingness to keep learning helped him grow far beyond his original expectations for the internship.

Supporting the Next Generation of Ocean Innovators

At Open Ocean Robotics, internships are more than temporary work experiences.

Students become part of multidisciplinary teams working on technologies that are actively deployed to support maritime security, environmental monitoring, scientific research, and autonomous operations around the world.

Carlos' journey as an intern reflects Open Ocean Robotics' commitment to developing the next generation of talent within Canada's growing blue economy.

"I've learned so much across the board," he says. "The biggest reason is the people. Everyone is incredibly knowledgeable, incredibly supportive, and always willing to help. It's been an awesome experience."

As Open Ocean Robotics continues advancing technologies like Enhanced Horizon, investments in students, interns, and early-career professionals remain an important part of building the future of ocean technology.

If you're passionate about artificial intelligence, robotics, software development, or creating technology that helps better understand and protect our oceans, we'd love to hear from you.

Explore current career opportunities at Open Ocean Robotics and help shape the future of persistent autonomous ocean intelligence.