Specifications

Processor

Qualcomm QCS6490 (8x Kryo 670 CPU, Adreno GPU, Hexagon NPU)

AI Performance

Up to 12 TOPS within a sub-10 W power envelope

Form Factor

100 x 60 mm compact board

Board Weight

~50 g without heatsink

Power Input

5-12 V DC

Thermal Design

Passive / fanless

Operating Temperature

-20 °C to 70 °C

Vibration Resistance

3.5 Grms

Protection

Industrial-grade ESD protection; lockable connectors

Camera Interfaces

5x 4-lane MIPI-CSI (RGB, monochrome, stereo, high-resolution)

Wired Networking

2.5 GbE; 2x USB 3.2 Gen 2 Type-C

Fieldbus / I/O

CAN FD, UART, I2C, GPIO for LiDAR, depth cameras, IMUs, flight controllers, ESCs

Wireless

Wi-Fi 6E, Bluetooth, 4G/5G via M.2, onboard Nano-SIM

OS

Ubuntu 24.04 LTS

Robotics Software

Advantech Robotic Suite for Drone; ROS 2 with PX4 and ArduPilot

Flight Stack

MAVLink, MAVSDK and MAVROS integration

Autonomy Workloads

Object detection/tracking, VIO, SLAM, obstacle awareness, GNSS-denied navigation

Lifecycle

Long-term product lifecycle support

Overview

The Advantech ASR-D501 is a compact AI companion and mission computer built for autonomous unmanned aerial vehicles (UAVs). Powered by the Qualcomm QCS6490, it delivers up to 12 TOPS within a sub-10 W envelope — a power budget that matters when every watt and gram of payload affects flight time.

Purpose-built for size-, weight- and power-constrained airborne platforms, the ASR-D501 provides a unified compute base for real-time inference, visual perception, localisation, sensor fusion and mission processing. It exposes five 4-lane MIPI-CSI camera interfaces, 2.5GbE, two USB 3.2 Gen 2 Type-C ports and onboard CAN FD, UART, I2C and GPIO for LiDAR, depth cameras, IMUs and flight controllers, with Wi-Fi 6E and 4G/5G for BVLOS links. The Advantech Robotic Suite for Drone pre-integrates ROS 2 with PX4 and ArduPilot via MAVLink/MAVSDK/MAVROS.

Key Benefits

Airborne autonomy without a desktop GPU: the ASR-D501 keeps inference and sensor fusion onboard at under 10 W, freeing the flight controller for stabilisation and control. Five camera inputs, lockable connectors and a -20 to 70 °C, 3.5 Grms rating make it production-ready for UAVs rather than a bench prototype.

Applications

Autonomous drone perception and navigation, VIO and SLAM, obstacle awareness, GNSS-denied navigation, commercial UAV payloads, and robotics sensor fusion.

Request a Quote — ADVANTECH ASR-D501 — QUALCOMM QCS6490 AI COMPANION COMPUTER FOR UAVS

QS Compute — global B2B supply of AI computing hardware, edge AI systems and accelerators. Volume pricing, 15-day sample lead time.

Get Your Quote →

Related Products

Qualcomm QCS6490 Edge AI Box Qualcomm QCS8550 Edge AI Box Geniatech Edge AI Box Series Khadas Edge2 / Edge-V Series