Skip to main navigation Skip to main content Skip to page footer
RepliCar · Compute and software

Backend

NVIDIA Jetson AGX Orin · Wellenzahl carrier board

Receive, process, deliver — on one compact unit.

The RepliCar Backend terminates up to eight radar links, supplies the modules through the same cables, runs the complete processing chain and provides results and raw data over standard interfaces. It combines an NVIDIA Jetson AGX Orin module with a carrier board that Wellenzahl developed specifically for RepliCar.

8
radar links incl. supply
10Gbit/s
Ethernet to your systems
6
angle-estimation methods
ROS 2
· TCP · UDP outputs
RepliCar carrier board: connector for the NVIDIA Jetson AGX Orin module, two A-PHY deserializers and eight coax inputs for the radar modules.
Wellenzahl carrier board · module sideEight coax inputs
01 · Carrier board

Developed for RepliCar. Nothing else required.

Instead of a development kit with add-on cards, the Jetson AGX Orin module sits on a carrier board designed by Wellenzahl. It brings everything a radar system needs: the inputs for the radar links, their supply, storage for recordings and fast interfaces to your infrastructure.

WELLENZAHL CARRIER BOARDRADAR LINKS1234A-PHY deserializer4 radar linksraw data in,control back channel outCSI-25678A-PHY deserializer4 radar linksraw data in,control back channel outCSI-2Port power12 V per link · switchable · current-limitedNVIDIA Jetson AGX Orincamera interfaceArm CPU · Ampere GPULPDDR5 memoryradar driver (V4L2)processing chainweb UI · profilesrecording · replayPower management12 V DC or USB-C PD input10 Gigabit Ethernetresults · recordingsM.2 NVMe slotlong recordings2 × USB-CperipheralsDisplayPortlocal monitormicroSDstorage
Functional block diagram · simplified
Interface side of the RepliCar carrier board: 12 V supply input, DisplayPort, 10 Gigabit Ethernet, two USB-C ports and M.2 slot.
Interface side · 12 V input, DisplayPort, 10 Gigabit Ethernet, 2 × USB-C, M.2 slot

Eight radar inputs

Two A-PHY deserializers receive up to eight radar links. Raw data flows directly into the camera interface of the Jetson; control commands travel back to the modules on the same cables.

Supply over the radar cable

Every input supplies its module with 12 V over the coax cable — individually switchable, current-limited and with a status indicator per port.

Interfaces for your systems

10 Gigabit Ethernet for results and recordings, an M.2 NVMe slot for long recordings, USB-C, DisplayPort and microSD.

02 · Data path

From the antenna to the point cloud — without detours.

Raw data reaches the memory of the Jetson without being compressed, packed into network packets or copied by the CPU. That keeps latency low and leaves the computing power for the signal processing.

01 · Radar module

Digitise

The transceivers sample the echoes with 12 bit; both streams are merged into one.

02 · Coax link

Transport

MIPI A-PHY over up to 15 m of coax, uncompressed.

03 · Carrier board

Receive

The deserializers hand the streams to the camera interface of the Jetson.

04 · Memory

Buffer

Radar frames arrive as standard V4L2 video frames — zero-copy.

05 · Processing

Compute

Range-Doppler, detection, angle estimation, clustering, tracking — CPU or GPU.

06 · Outputs

Deliver

Point clouds and tracks via ROS 2, TCP or UDP; raw data in parallel.

16
V4L2 streams at full scale · two per subarray
36.8Gbit/s
peak payload of eight links
≈2.1GiB/s
raw data at 30 Hz · Full Aperture
≈540MiB/s
raw data at 30 Hz · Compact Cascade

Why C-PHY towards the Jetson? C-PHY carries about 2.28 bits per symbol on three-wire trios — roughly twice the bandwidth per pin of D-PHY — so the complete raw-data stream fits through the camera interface without running out of lanes.

03 · Signal processing

Six stages. Every one switchable at runtime.

The complete chain runs in Rust on the Jetson. The CPU is sufficient for the Compact Cascade; CUDA takes over at full scale. Every stage can be switched during operation, from fast to high-resolution.

01 · MIMO

MIMO front end

The transceiver array delivers one coherent data cube per frame — chirps, channels and samples, phase-locked across all subarrays.

FMCW chirp · data cube
02 · 2D-FFT

Range-Doppler

A two-dimensional FFT per channel produces range and velocity maps — the basis for detection and angle estimation.

Range / velocity map
03 · CFAR

Detection

An adaptive CFAR threshold finds targets reliably, even under varying noise and clutter.

Adaptive threshold
04 · DoA

Angle estimation

Selectable methods — from fast Bartlett to high-resolution Capon and MUSIC, plus the sparse methods OMP, ISTA and IHT — compute azimuth and elevation.

Bartlett
05 · DBSCAN

Point cloud and clustering

Detections are grouped into 3D objects, including heading and extent.

Cluster · bounding box
06 · IMM

Multi-target tracking

An IMM tracker assigns stable IDs over time and smooths position and velocity.

Tracks · stable IDs
04 · Software

Four layers, from the driver to the browser.

Driver

V4L2 radar driver

Linux kernel

Presents radar frames as standard video frames with 12-bit raw data. Any V4L2-capable tool can capture them; one software image serves one to eight subarrays.

Control and capture

Radar control

radar-cli

Controls all connected modules, switches measurement profiles at runtime, records and replays raw data and serves the browser-based radar UI.

Processing

Signal processing

radar-processor

Range-Doppler processing, CFAR detection, six angle-estimation methods — Bartlett, Capon, MUSIC, OMP, ISTA, IHT — DBSCAN clustering and IMM tracking, with a CUDA backend for large apertures.

Outputs

Results and raw data

ROS 2 · TCP · UDP · V4L2

Tracked point clouds and 3D bounding boxes for your application, ready for common tools such as Foxglove and RViz. Raw data stays accessible in parallel.

Browser-based RepliCar radar UI showing live range profiles of all channels.
Browser-based radar UI · configuration, live view and recording, directly on the backend
05 · Integration

Three ways to work with RepliCar data.

Use the results

Point clouds and objects

Subscribe to tracked point clouds and 3D bounding boxes via ROS 2, or receive them over TCP or UDP in your own software.

Choose the method

Switch processing at runtime

Select between the built-in processing methods during operation — for example from fast Bartlett to high-resolution Capon or MUSIC beamforming.

Work on raw data

Live or recorded

Access the 12-bit raw data of every transceiver live, or record it and replay it into the processing chain as often as needed.

One detection, as an example

Typical content of a detection: position in 3D, radial velocity, signal strength, cluster and track assignment and a time stamp. The interface description documents fields, coordinate frame, units and time base.

// illustrative field names
{
  "frame": 18342, "t_ns": 1739871234567890123,
  "x_m": 12.41, "y_m": -1.87, "z_m": 0.36,
  "v_r_mps": -4.12, "snr_db": 21.7,
  "cluster_id": 3, "track_id": 17
}
06 · Measurement profiles

Managed on the backend. Extended as you need.

Profiles define how the radar measures within the 76–81 GHz band: sweep bandwidth, chirp sequence, transmit scheme and frame rate. The backend distributes them to all connected modules and switches them during operation.

Included

Tested profiles

Eight tested profiles from high range resolution to long range, ready for use on delivery.

Your own

Add profiles yourself

Define profiles for your scene and switch to them at runtime, without restarting the system.

Engineering service

Developed by Wellenzahl

On request, we develop and validate profiles for specific targets, ranges and velocities.

07 · Performance

Measured on the backend.

<30ms
end-to-end latency · Full Aperture
≈95W
complete system · 130 W peak
≈37Hz
3,072 channels · GPU
≈100Hz
192 channels · CPU
ConfigurationComputeProcessing time per frameProcessing rate
Compact Cascade · 192 channelsCPU9–11 ms91–111 frames/s
Full Aperture · 3,072 channelsCPU39 ms25 frames/s
Full Aperture · 3,072 channelsGPU (CUDA)≈27 ms≈37 frames/s

NVIDIA Jetson AGX Orin 64 GB. Processing time covers the computation per frame. End-to-end latency is measured from capture by the radar to the processed output; power consumption refers to the complete Full Aperture system including the backend.

08 · Technical data

At a glance.

Compute
Compute moduleNVIDIA Jetson AGX Orin (benchmarks with 64 GB)
Operating systemLinux · NVIDIA JetPack 6.2
Radar interfaces
Radar links8 × MIPI A-PHY over coax
Module supply12 V per link · switchable · current-limited · status indicator
ControlVia the back channel of each link
Host interfaces
Network10 Gigabit Ethernet (10GBASE-T, RJ45)
StorageM.2 NVMe slot (Key M) · microSD
Further interfaces2 × USB-C · DisplayPort
Software and outputs
OutputsTracked point clouds and 3D bounding boxes via ROS 2, TCP, UDP
Raw dataV4L2 video devices, 12-bit · recording and replay
OperationBrowser-based radar UI · profile switching at runtime
Supply
Supply input12 V DC or USB-C Power Delivery
Power consumption≈95 W typical, 130 W peak for the complete Full Aperture system · backend alone: see datasheet

Your algorithms on our data?

Request raw-data recordings, a live session on the system or a backend adapted to your interfaces.

WellenzahlЯEPLICAR · MODULAR IMAGING RADAR BY WELLENZAHL