AidLux is an AIoT development platform that runs a native Linux environment with GUI and AI capabilities on Android and HarmonyOS devices.
The platform addresses the challenge of deploying AI models and running development tools on mobile and edge devices by creating a unified environment where Android and Linux coexist on the same kernel without requiring virtualization or system reboots. It provides heterogeneous compute scheduling across CPU, GPU, and NPU through its AidLite SDK, with optimizations for Qualcomm chipsets and support for quantized inference using INT4, INT8, and FP16 formats. The architecture enables seamless access to Android hardware drivers like cameras and sensors while simultaneously providing Linux software stacks for AI development.
Developers should choose this tool if they need to deploy large language models, vision models, or embodied AI applications directly on mobile devices or edge hardware. It suits robotics development through native ROS and ROS2 integration, supports nearly five hundred edge-optimized models including Qwen, Phi, and Deepseek variants, and includes specialized tools like AidCode for interactive Python development and AidTerminal with touch-optimized controls. The platform is particularly valuable for teams wanting to avoid the complexity of cross-compilation or traditional mobile development constraints while maintaining access to both Android and Linux ecosystems.
Development activity shows consistent feature expansion with recent updates introducing Ubuntu-desktop as the default environment, extended hardware support for newer Snapdragon chipsets, and integration of updated Qualcomm Neural Network inference libraries. The project maintains active refinement of core infrastructure including WiFi management, Linux installation workflows, and desktop interface usability. Regular updates address both capability expansion and optimization of the underlying heterogeneous compute scheduling system.
