Deep Dive: aily Blockly - An AI-Native Hardware Development IDE
This is an English summary and translation of a Chinese forum research report on aily Blockly.
Key points
- Positioning: aily Blockly bills itself as "the world's first AI-native hardware development environment," aiming to blur the line between professional and non-professional development, ultimately enabling natural-language programming. It is explicitly a *productivity tool* with an industrial-grade software architecture, not an educational toy.
- Tech stack: Electron + Angular + Node.js + Google Blockly; licensed under GPL v3. GitHub: https://github.com/ailyProject/aily-blockly — Official site: https://aily.pro — Status: Alpha.
- Desktop app (Electron/Angular) with three main UI components: visual Blockly editor, serial debugging tool, and an AI chat agent.
- Project manager (npm-based): each project manages its own board package and library dependencies (e.g.,
esp32@2.0.14in anaily-project.json), avoiding the global version conflicts of the Arduino IDE. - Lightning compilation (edge-cloud collaboration): lightweight local syntax checking plus pre-configured cloud build environments with caching and parallelism — claimed speedup from ~1 hour local to ~1 minute, a 100%+ improvement.
- AI Agent layer: three agents — Nuwa (board configuration), Luban (library conversion), and a project generator for automated project setup.
- AI project generation (small-scale testing): natural language → project architecture, pinout diagram, and code.
- AI code generation (small-scale testing): writes programs from requirements.
- AI library conversion (small-scale testing): converts C/C++ Arduino libraries into Blockly blocks automatically (analyze headers/implementations → generate block definitions and generators → publish to library manager). Goal: the platform with the most Blockly libraries.
- AI board configuration (in progress): generate board configs from markdown documentation.
- Serial tool: data visualization charts, multiple encodings, logging, real-time debugging.
- Multi-repo structure: main IDE, boards, libraries, compiler, and tools repositories.
- Most block libraries are currently AI-generated and require community validation.
- IDE is fully free and open source (GPL v3). Potential monetization (speculative): cloud compilation, AI tokens (report cites 46M+ tokens/24h consumption), enterprise editions, hardware partnerships. Current sponsors include Tianwei Electronics, MUSHROOM Cloud makerspace, and Yixuehui.
- Done (Alpha): project management, library manager, serial tool, lightning compilation phase 1.
- Planned: hardware simulation (TBD), MicroPython mode (added, no libraries yet), AI board config and AI block generation (targeted 2025-04), lightning compilation phase 2.
- GitHub: https://github.com/ailyProject/aily-blockly
- Official site: https://aily.pro
- Boards repo: https://gitee.com/coloz/aily-blockly-boards
- Blog: https://www.cnblogs.com/coloz999/p/19165688
- Juejin: https://juejin.cn/post/7564661612319211535
Architecture
Core features
Supported hardware
Arduino AVR (Uno, Mega), ESP32 (S2/S3/C3), STM32 (F1/F4/L4), RP2040/RP2350, Renesas RA4/RA6, Nordic nRF52/nRF53. Board configs live at https://gitee.com/coloz/aily-blockly-boards; new boards can be added by hand-written JSON or AI-generated from docs.
Ecosystem and business model
Competitive comparison
| Dimension | aily Blockly | Mixly | Arduino IDE | PlatformIO | |---|---|---|---|---| | Target users | All levels | Education/beginners | Beginner-mid | Professionals | | Programming | Visual + AI | Visual + code | Code | Code | | Engineering | npm-based | Legacy | Simple | Mature | | AI capability | Very strong | Minimal | Minimal | Limited | | Compile speed | Very fast | Slow | Moderate | Fast |
Key differentiators: AI-native design from day one, npm dependency isolation and reproducible builds, lightning compilation, and theoretically unlimited library support via AI conversion.
Roadmap
Assessment
Strengths: genuine AI-native design; solves embedded version-management pain; major compile speed gains; open source; scalable library ecosystem.
Challenges: Alpha-stage software (not recommended for production); AI-generated libraries need verification; heavy reliance on LLMs (cost/stability uncertain); smaller community than Arduino.
Best use cases: rapid prototyping, education, AI-assisted fast development, and porting existing C/C++ libraries to visual programming. If AI features mature, the report suggests it could become a "Cursor for embedded development."