Why This Matters

Coastal monitoring programs are often constrained by labor, coverage, and detection timing. We aim to complement existing methods by improving spatial visibility and helping teams prioritize intervention.

About

I'm Michael Walsh, founder of Coastal Research Robotics. My background is in tech and hardware, with undergraduate work in chemistry and biology — a combination that turned out to be useful in ways I didn't expect.

In 2025 I started crabbing for Dungeness crab in Bellingham Bay. Around the same time, a piece in Cascadia Daily caught my attention: European green crab had established populations in local coastal waters and were actively threatening Dungeness crab habitat. I was also, at that point, deep into a computer vision project at work.

The overlap was hard to ignore. Green crab are a documented predator pressure on juvenile Dungeness and a known disruptor of the eelgrass habitat Dungeness depend on. Existing monitoring programs rely on trap surveys — labor-intensive, spatially limited, and slow to detect new incursions. I had a background in data analytics and was already building CV pipelines. An underwater ROV with a species classifier felt like a direct response to a real and specific problem.

That's what Coastal Research Robotics is: a bet that the right combination of field robotics, computer vision, and ecological data can give resource managers something they don't currently have — early, spatially explicit detection at the scale the problem requires.