We introduce a large-scale, fully integrated 4D imaging sensor built on a coherent FMCW LiDAR focal plane array with 352 by 176 pixels, delivering real-time 4D imaging of dynamic scenes including distance, velocity, and intensity.

The FPA is monolithically integrated on a silicon photonics platform with CMOS electronics, featuring 16 optical channels, a two-stage thermo-optic switch network, and a dual-path coherent detection scheme that uses heterodyne reception to suppress noise and enable precise measurements.

A key architectural choice is the single-coherent-pixel design, where each pixel both emits and collects light via a balanced photodiode pair and a local oscillator, forming a monostatic arrangement that eliminates optical cross-coupling and simplifies alignment.

The array is organized into blocks of 8-pixel rows with a 1:8 optical splitter per row, enabling 128-pixel parallel operation and achieving real-time 4D imaging at up to 15 frames per second with optimized readout.

The system offers a radial range of 65 meters, angular resolution of 0.06 degrees, and operates with very low per-point energy (about 46 nanojoules) and modest on-target optical power per pixel (roughly 178 microwatts), indicating scalable low-power long-range 3D imaging.

An experimental setup uses a 1310 nm FMCW source with IQ modulation, delivering light through 16 fibers to the FPA, with off-chip processing to extract per-pixel distance, velocity, and amplitude; off-the-shelf SWIR lenses determined the field of view and resolution.

Future work aims to extend range beyond 200 meters, boost SNR by approximately 5.6 dB through pixel design refinements, mitigate silicon nonlinear losses with new materials, and relocate second-stage switches outside the FPA to improve pixel placement and far-field uniformity.

The work situates itself within a broad literature base, citing foundational and recent studies from IEEE and Nature to anchor the technology in ongoing photonics and quantum electronics research.

Positioned as a CMOS camera-like platform for multidimensional imaging, the FPA promises cost reductions, compact form, eye-safety, and modular scalability suitable for markets needing high-resolution, real-time 4D sensing.

Licensing notes indicate CC BY 4.0 for Expert Opinion content and CC BY-NC-ND 4.0 for Figure 1, with a download link provided for references.

A new affordable, low-power scalable sensor based on integrated silicon photonics maps position and velocity in dynamically evolving 3D scenes, effectively enabling 4D imaging.

Key references cited include foundational and recent works from Behroozpour, Shekhar, Zhang, Poulton, and Rogers, establishing the scholarly lineage of the approach.