OmniPreSense Internship

I deployed an OPS243 coverage calculator, explored real-road 2D and 3D views, and submitted OpenFlight internal-trigger PR #221.

Role: Engineering intern
Year: 2026
Discipline: Embedded sensing, radar, web engineering

Scroll to case study

01 · Production deployed

OPS243 field-of-view calculator

I turned flat, straight-road radar placement geometry into a public tool for planning OPS243 sensor coverage before deployment.

Problem

A radar installer needs to understand lane coverage before mounting an OPS243 sensor. The original calculations needed a clear field workflow for a flat, straight road.

My implementation

As an engineering intern, I carried MATLAB geometry and lane-intersection calculations into the web calculator.

I implemented metric and imperial inputs for mounting height, vertical and horizontal angle, sensor-to-road-center distance, road side, and lane width. The tool returns field-of-view dimensions and center, per-lane travel distance, and road-view and full-FOV diagrams.

Blender studio render of the OPS243 radar evaluation board in a three-quarter view — Assembled OPS243 board render with its radar module, antenna traces, and evaluation hardware.

Overhead Blender render of the OPS243 antenna array and evaluation PCB — Overhead view of the OPS243 antenna array and evaluation PCB.

Public deployment: Open OPS243 field-of-view calculator
Primary result: Radar coverage and lane travel-distance planning

02 · Prototype / active R&D

Real-road 2D mapping

I explored how to place radar field of view against actual roads instead of idealized straight geometry.

Problem

A flat, straight-road diagram cannot show where a sensor sits on a real roadway. Curves and lane geometry need to stay connected to a geographic map.

My exploration

I integrated OpenStreetMap-related roadway geometry and worked toward plotting radar field of view against actual roads, including curves and lanes.

I explored interactive sensor positioning and road-based placement workflows on mapped roads.

03 · Early R&D / exploration

3D / perspective FOV visualization

I investigated how radar coverage could be understood in a 3D or 360-degree environmental context.

Problem

A flat FOV diagram cannot communicate perspective, surrounding structures, or a real deployment viewpoint. A field team could benefit from seeing coverage in context.

What I explored

I researched Google Street View and 360 imagery concepts, local 3D reconstruction, Gaussian splatting, and perspective-correct radar visualization.

I kept this work at the early R&D stage while the production calculator remained a flat, straight-road tool.

04 · Submitted / open PR #221

Hardware-triggered OpenFlight capture

I built the OPS243 capture path for golf-swing testing with OpenFlight processing.

Problem

The existing rolling-buffer capture relied on an external SEN-14262 sound-trigger edge. I needed an opt-in OPS243 internal hardware-trigger strategy so the radar could make the speed-trigger decision and dump rolling I/Q history without that external module, while preserving the sound-trigger default.

My implementation

I configured and validated OPS243 firmware and trigger settings, waited for the board-generated dump, parsed the capture, and re-armed the workflow before passing parsed radar data into OpenFlight processing.

I validated the implementation on real OPS243 hardware running firmware v1.3.1 and tested it against the latest OpenFlight release available during validation, including trigger, signal, capture, and parser behavior.

Upstream contribution: View OpenFlight PR #221