Sounds like a solid challenge. Keep the sensor array minimal—maybe just a temperature probe, a pH sensor, and a tiny light meter. Pack them in a single, flat module so it doesn’t snag on reef structures. Use a low‑power microcontroller with a Bluetooth LE link; the data can scroll out to a phone or a dock in the same pocket as the device. A single, matte‑black case with a subtle matte‑white line for the cable will give it that sleek feel. Make sure the battery is a small, replaceable cell that fits the same shape; that way you’re not adding bulk. Remember, precision is a tool, not the goal. If you can get the right data without making the user think about it, you’ve hit the sweet spot. Good luck—let me know if you need a CAD sketch to start the iterations.

A 3.7 V Li‑Po with a capacity of around 100 mAh can keep the low‑power sensors running for a week or two if you keep the MCU in deep‑sleep between readings. If you expect sudden bursts—like a temperature spike or a pH jump—a small supercapacitor in parallel can absorb those spikes without draining the battery. So, Li‑Po plus a tiny capacitor is a practical combo. For the layout, picture a flat, 30 mm long strip. On the outer edge you place the temperature probe, flush with the surface so it measures the water directly. Next to it, a small pH electrode sits just behind a thin, non‑reactive spacer so it only contacts the water, not the coral. The light sensor sits on the opposite side of the strip, angled slightly to capture ambient light without shading the other sensors. All three connect to a single PCB that also hosts the microcontroller and the BLE module. The battery sits at the far end of the strip, with a minimal connector to keep the shape low‑profile. That keeps everything balanced and keeps the footprint as small as possible.

For the pH probe, I’d embed it in a tiny, hermetically sealed glass capsule that sits flush with the strip’s surface. A thin, flexible spacer—just a few millimetres thick—keeps the electrode away from the bulk of the device so it never touches the coral. The capsule is sealed with a silicone gasket, which resists biofouling and keeps the electrolyte stable. The probe is temperature‑compensated; the microcontroller applies a standard correction formula based on the thermometer reading. The light sensor stays calibrated by periodic exposure to a calibrated LED source that emits a known lux level. I’d program the device to log a “dark” reading when the strip is covered, then use the LED to provide a reference reading. Over time, the algorithm adjusts the sensor’s output curve to match the reference, correcting drift. If you need a quick layout, imagine the strip as a 30 mm long line: temperature probe on the left edge, pH capsule in the middle, light sensor on the right edge, all connected to a single board. The battery sits at the far end. I can send you the CAD file once it’s ready for review.