Tactical SAR Hexacopter - integrating radar + thermal with Pixhawk 6C

General
1

Hi everyone,

I’m designing a tactical Search and Rescue hexacopter to detect survivors under rubble or behind walls, combining a 60GHz UWB radar with a thermal camera and a depth camera, processed onboard and streamed live to rescue teams.

FRAME & FLIGHT CONTROL

- Frame: Tarot X6 folding hexacopter, 10.6"

- Flight controller: Pixhawk 6C

- Companion computer: Jetson Orin Nano 8GB

MOTORS & POWER

- Motors: T-Motor MN4014 330KV (x6)

- Props: 18x6.1" carbon fiber

- ESCs: FLAME 60A 6S (x6)

- Power distribution: Matek FCHUB-12S

- Battery: Tattu 23000mAh 25C 6S LiPo

SENSORS

- Radar: Acconeer XM122 (60GHz UWB, penetrates rubble/walls to detect breathing/heartbeat)

- Thermal camera: FLIR Lepton 3.5

- Depth camera: Intel RealSense D435i

COMMS

- Primary link: HereLink Blue

- Backup telemetry: RFD900x

Estimated thrust-to-weight ratio: 2.25:1, AUW ~8kg, estimated flight time ~32 min.

Still in the design phase, nothing purchased yet. I’d really appreciate feedback on:

1. Integrating radar/thermal sensors with Pixhawk/ArduPilot (companion computer setup, MAVLink integration)

2. Payload weight management for SAR builds

3. Any lessons learned from similar projects

Thanks in advance!

2

XM122 seems like an interesting sensor, but i think your AI of choice has oversold you on its capabilities.

For example the data sheet states a range of only 7m and makes no mention of penetrating rubble, unless you’re intending to detect the motion of a thin layer of rubble. The use case for breathing detection seems to be purely in a medical case where a known targets chest deflection is measured

3

Thanks Alex, this is a really helpful correction — I appreciate you pointing to the actual datasheet numbers instead of letting the AI’s framing stand uncorrected.

You’re right that I had the wrong picture: I was assuming rubble-penetration and breathing detection through debris, but the ~7m range and the medical chest-deflection use case you mentioned make clear that’s not what this sensor was built for.

For my SAR hexacopter project, my remaining sensor stack is a FLIR Lepton 3.5 thermal camera and an Intel RealSense D435i depth camera. Given the XM122’s real capabilities, do you think it still adds value as a short-range presence/motion sensor for partially-exposed victims, or would I be better off dropping it and reallocating that budget elsewhere (e.g. battery redundancy or prop guards)?