Hi Bristlemouth community — this is my first post as a Pioneer recipient, and I wanted to kick things off with a big thank you and a look at what I’m planning to build.
A bit of background: I spent the last ~7 years on the hardware engineering team at Sofar, designing many of these BM parts and pieces. Now I get to flip roles and become an end-user of the tech I love — building my own BM-powered modules to help grow the ecosystem. Some of you may have seen my earlier posts where I developed an open-source BM + Raspberry Pi camera module for early ocean testing. Those source files are still live, and that positive experience is what pushed me to dive in and start a company — Nereus Vision — with the mission of building low-cost, persistent marine cameras so others can see where they previously couldn’t.
I’ll be tracking the whole journey here on the forum, but first: a genuine thank you to Sofar and the broader BM support organizations. It was great to visit the Sofar office to pick up the Spotter + Smart Mooring in person and see so many friendly, familiar faces — I’m really looking forward to collaborating with everyone on this one.
What I’m building
My goals for this project:
- Develop a commercial product around a BM-powered camera
- Support both above- and below-water applications
- Operate reliably in power- and bandwidth-constrained environments
- Run inference models onboard and send alerts or still images
- Keep it an open platform, so others can deploy their own computer-vision models to run at the edge
Where I’d like this to go
- Self-clearing wiper that detects a dirty viewport and keeps it clean of biofilm underwater
- Onboard video recording saved to local memory
- Video streaming down the road (2–4 Mbps to shore) once telemetry options can support the bandwidth
- Add-on light modules for deeper, darker depths
- Self-contained timelapse mode — deploy for months at a time on batteries
- Remote control via the Sofar command API, so you can reconfigure the camera in the field without recovering and redeploying
- Data streaming to a custom website (already up and running) to monitor the camera, get alerts, and run cloud computer-vision processing on the compressed images you receive
How the Pioneer kit helps
- Bench testing — use the dev kit to talk to different camera sensors and evaluate image size/quality and power
- Pool testing — run the hardware in a controlled underwater environment
- Field testing + collaboration — ideally partnering with a group that has permitted sites on the CA coast to push the hardware’s limits
Rough timeline
- Bench testing — OpenMV AE3 vs. N6 vs. Pi Zero with a 12MP camera sensor
- Pool testing — build one “mega camera” running all three at once, timelapsing underwater reference cards in changing light, benchmarking image quality, AprilTag detection, and object detection/counting in a controlled setting
- Field testing — prolonged real-world deployments to see how pool results translate to the wild
- Refine + early production — tighten the design, build a small run of validation units, and get them into early customers’ hands
Here is a draft timeline for how I’m planning the rollout. Nothing is set in stone, I’m using this as a general way to track progress and milestones.
Stay tuned — in a follow-up post I’ll put out a short survey to hear about your specific use cases and applications for marine cameras. I’d genuinely value learning what you’d want from a platform like this.
PS - that is Oakley in the above image, you’ll be seeing more of her in the future!