🧭 IgnitionAI - Project Roadmap

This document outlines the phased development of IgnitionAI — a modular, browser-friendly framework for intelligent agent simulation and reinforcement learning.

---

✅ Phase 1 — Core Logic (MVP)

⚙️ Goal: Run agent-environment logic headlessly (no UI)

✅ Roadmap for "RL algo first" Phase A — @ignitionai/backend-tfjs only Implementing classic algorithms with TensorFlow.js

1. 🔁 Q-learning (tabular) – minimalist JS version without neural networks ✅ Implemented Q-Table agent with state/action lookup ✅ Added tests for basic functionality

2. 🧠 DQN – Deep Q-Network ✅ Implemented MLP simple input → hidden → output ✅ Added replay buffer with experience sampling ✅ Implemented target network with periodic updates ✅ Added epsilon-greedy exploration/exploitation ✅ Loss function based on TD error ✅ Unit tests with training validation

3. 🧘‍♂️ PPO – Policy Gradient ✅ Created initial PPO agent skeleton

• Implement Actor-Critic model
• Implement episode-based training
• Add policy and value loss functions

---

✅ Phase 1.5 — Backend Infrastructure

🧰 Goal: Create robust, multi-environment backend support

✅ Created modular monorepo structure ✅ Implemented robust backend selection system ✅ Added support for all major TensorFlow.js backends:

• WebGPU (experimental)
• WebGL
• CPU
• WASM ✅ Added helper utilities for backend detection and info ✅ Added comprehensive model management system:
• IndexedDB local storage
• Hugging Face Hub integration with authentication
• Automatic model serialization/deserialization
• Checkpoint system with:
  • Regular checkpoints (step-based)
  • Best model checkpoints
  • Automatic retry with exponential backoff
• Model versioning and metadata ✅ Added robust error handling and logging ✅ Comprehensive unit tests and integration tests

---

🚀 Phase 2 — R3F Visualisation

🎮 Goal: Make the agent & target visible in a 3D scene

✅ @ignitionai/r3f: add AgentMesh, TargetMesh, useAgent ✅ @ignitionai/demo-target-chasing: setup Vite + R3F scene ✅ Add training monitoring and auto-stop functionality ✅ Display step count and reward in the UI ✅ Implement real-time model updates

• Add training controls and visualization
• Optimize performance for longer training sessions
• Add ability to save/load models from the UI

---

✅ Phase 3 — TFJS Backend (Training & Inference)

🧠 Goal: Train and run a model directly in the browser

✅ @ignitionai/backend-tfjs: built simple MLP model with configurable layers ✅ Implemented train() and predict() APIs via DQN agent ✅ Added model serialization with save() and load() ✅ Added support for Hugging Face Hub integration ✅ Created streamlined Agent class interface ✅ Added comprehensive training utilities:

• Progress tracking
• Performance metrics
• Model checkpointing
• Training visualization ✅ Implemented browser-based training with Three.js visualization ✅ Added automatic checkpoint saving for best models

---

🚀 Phase 4 — ONNX Runtime Backend (Inference-only)

⚡ Goal: Run optimized pre-trained models in production

✅ Created initial package structure for ONNX backend

• Implement ONNX Runtime Web integration
• Add .onnx model loading and inference
• Create InferenceBackend wrapper
• Add model conversion utilities (TFJS → ONNX)

---

🚀 Phase 5 — Advanced Environments

🌍 Goal: Create more complex environments for agent training

• Implement grid-based environments (maze, pathfinding)
• Add physics-based environments (pendulum, cartpole)
• Create multi-agent environments
• Add environment customization tools
• Implement environment visualization tools

---

🚀 Phase 6 — Advanced Algorithms

🧠 Goal: Implement more sophisticated RL algorithms

• Implement DDPG (Deep Deterministic Policy Gradient)
• Add SAC (Soft Actor-Critic)
• Implement A2C (Advantage Actor-Critic)
• Add support for custom algorithm implementations
• Create algorithm comparison tools

---

🚀 Phase 7 — Deployment & Production

🚢 Goal: Make the framework production-ready

• Add comprehensive documentation
• Create example applications
• Implement CI/CD pipeline
• Add performance optimization tools
• Create deployment guides
• Add monitoring and analytics

---

Built with ❤️ by Salim (@IgnitionAI)