diff --git a/docs/TROUBLESHOOTING.md b/docs/TROUBLESHOOTING.md
new file mode 100644
index 0000000..241edc7
--- /dev/null
+++ b/docs/TROUBLESHOOTING.md
@@ -0,0 +1,306 @@
+# Troubleshooting
+
+Common errors and fixes when deploying Reflex VLA on edge devices, cloud GPUs, ROS2 robots, and drones.
+
+---
+
+## CUDA & GPU Errors
+
+### `libcudnn version mismatch`
+
+```
+Could not load library libcudnn_ops_infer.so.8. Error: libcudnn_ops_infer.so.8: cannot open shared object file
+```
+
+**Cause:** Your cuDNN version doesn't match what ONNX Runtime was compiled against.
+
+**Fix:**
+```bash
+# Check your versions
+python3 -c "import torch; print(torch.version.cuda)"
+python3 -c "import onnxruntime; print(onnxruntime.get_available_providers())"
+
+# Install matching cuDNN (example for CUDA 12.x)
+pip install onnxruntime-gpu==1.18.0  # matches CUDA 12.x + cuDNN 8.9
+```
+
+**Jetson fix:** On JetPack 6.0+, use the NVIDIA-provided ORT wheel:
+```bash
+pip install onnxruntime-gpu --extra-index-url https://download.onnxruntime.ai/onnxruntime_stable_jetpack60.html
+```
+
+---
+
+### `NVIDIA driver too old for CUDA runtime`
+
+```
+CUDA driver version is insufficient for CUDA runtime version
+```
+
+**Cause:** Your GPU driver is older than the CUDA toolkit Reflex was built against.
+
+**Fix:**
+```bash
+# Check driver version
+nvidia-smi
+
+# Minimum driver versions:
+# CUDA 12.0 → driver 525.60+
+# CUDA 12.2 → driver 535.54+
+# CUDA 12.4 → driver 550.54+
+
+# Update driver (Ubuntu)
+sudo apt install nvidia-driver-550
+sudo reboot
+```
+
+---
+
+### `Out of memory during graph capture`
+
+```
+RuntimeError: CUDA out of memory. Tried to allocate X MiB
+```
+
+**Cause:** The model + CUDA graph capture exceeds your GPU's VRAM.
+
+**Fix:**
+| GPU | Available VRAM | Max Model |
+|---|---|---|
+| Jetson Orin Nano | 8 GB | SmolVLA only |
+| Jetson AGX Orin | 32–64 GB | Pi0, Pi0.5 |
+| RTX 3060 | 12 GB | SmolVLA |
+| RTX 4090 | 24 GB | SmolVLA, Pi0 (LoRA) |
+| A100-40 | 40 GB | Pi0 full |
+| A100-80 / H100 | 80 GB | Pi0.5 distill |
+
+```bash
+# Reduce memory usage
+reflex export <model> --precision fp16     # use half precision
+reflex serve ./export/ --device cuda       # ensure GPU, not CPU fallback
+
+# Monitor VRAM
+watch -n1 nvidia-smi
+```
+
+---
+
+### `CUDAExecutionProvider not available`
+
+```
+[WARNING] CUDAExecutionProvider is not available. Falling back to CPUExecutionProvider.
+```
+
+**Cause:** ONNX Runtime was installed without CUDA support, or CUDA libraries aren't on `LD_LIBRARY_PATH`.
+
+**Fix:**
+```bash
+# Install GPU-enabled ORT
+pip install onnxruntime-gpu
+
+# Verify
+python3 -c "import onnxruntime as ort; print(ort.get_available_providers())"
+# Should show: ['CUDAExecutionProvider', 'CPUExecutionProvider']
+
+# If CUDA libs aren't found:
+export LD_LIBRARY_PATH=/usr/local/cuda/lib64:$LD_LIBRARY_PATH
+```
+
+---
+
+## Jetson-Specific Issues
+
+### `JetPack version not supported`
+
+**Minimum:** JetPack 6.0 (L4T R36.x) with CUDA 12.2+.
+
+```bash
+# Check JetPack version
+cat /etc/nv_tegra_release
+# or
+dpkg -l | grep nvidia-l4t-core
+```
+
+### `Thermal throttling during inference`
+
+**Symptoms:** Latency spikes after 5-10 minutes of continuous inference.
+
+**Fix:**
+```bash
+# Set max power mode
+sudo nvpmodel -m 0        # MAXN on Orin
+sudo jetson_clocks         # lock clocks to max
+
+# Add heatsink or active cooling for sustained workloads
+```
+
+> **For drone deployments:** Jetson modules in UAV enclosures have limited airflow. Use `--deadline-ms` to gracefully handle thermal throttle spikes rather than dropping frames.
+
+---
+
+## ROS2 Bridge Issues
+
+### `No state data received`
+
+```
+[ros2_bridge] unsupported image size/encoding
+```
+
+**Cause:** The bridge isn't receiving data on the expected topics.
+
+**Fix:**
+```bash
+# Check topics are publishing
+ros2 topic list
+ros2 topic echo /joint_states --once    # arms
+ros2 topic echo /mavros/imu/data --once # drones
+
+# Ensure topic names match your launch
+reflex ros2-serve ./export/ \
+  --state-topic /joint_states            # arms
+  # or
+  --state-topic /mavros/imu/data         # drones (MAVROS)
+```
+
+### `Empty state vector for drone topics`
+
+**Cause:** The bridge expects `msg.position` (from `JointState`) but receives `msg.orientation` (from `Imu`).
+
+**Fix:** Update to the latest `reflex-vla` — PR #121 adds auto-detection for IMU quaternion orientation. Or specify the correct topic type.
+
+### `ROS2 not found`
+
+```
+ModuleNotFoundError: No module named 'rclpy'
+```
+
+**Fix:**
+```bash
+# Source your ROS2 workspace
+source /opt/ros/humble/setup.bash
+source ~/ros2_ws/install/setup.bash
+
+# Then run
+reflex ros2-serve ./export/ ...
+```
+
+---
+
+## Drone / MAVROS Issues
+
+### `MAVROS not connecting to FCU`
+
+```
+[WARN] FCU: DeviceError: serial0: open: No such file or directory
+```
+
+**Fix:**
+```bash
+# Check serial port
+ls /dev/ttyACM* /dev/ttyUSB*
+
+# Launch MAVROS with correct port
+ros2 launch mavros mavros.launch.py fcu_url:=serial:///dev/ttyACM0:921600
+```
+
+### `IMU data not arriving`
+
+```bash
+# Verify MAVROS is publishing
+ros2 topic hz /mavros/imu/data
+# Should show 50-200 Hz
+
+# If 0 Hz, check FCU connection and MAVROS status
+ros2 topic echo /mavros/state --once
+```
+
+### `Action output not reaching flight controller`
+
+**Cause:** Reflex publishes to `/reflex/actions` as `Float32MultiArray`, but PX4/ArduPilot expects `mavros_msgs/AttitudeTarget`.
+
+**Fix:** You need a bridge node to convert. Example:
+```python
+# Convert reflex actions → MAVROS attitude target
+# action[0:3] = roll_rate, pitch_rate, yaw_rate
+# action[3]   = thrust (0-1)
+```
+
+---
+
+## Export & Validation Issues
+
+### `Opset 19 not supported`
+
+```
+onnxruntime.capi.onnxruntime_pybind11_state.InvalidGraph: Unsupported opset 19
+```
+
+**Fix:**
+```bash
+# Downgrade to opset 17
+reflex export <model> --opset 17
+
+# Or upgrade ORT
+pip install --upgrade onnxruntime-gpu
+```
+
+### `VERIFICATION.md says "Not yet verified"`
+
+**Cause:** Export completed but validation hasn't been run yet.
+
+**Fix:**
+```bash
+reflex validate ./reflex_export/
+# This populates the parity table in VERIFICATION.md
+```
+
+See [Understanding VERIFICATION.md](./understanding_verification.md) for interpreting the results.
+
+---
+
+## Model Registry Issues
+
+### `Unknown embodiment preset`
+
+```
+ValueError: Unknown embodiment preset 'myrobot'.
+```
+
+**Fix:**
+```bash
+# See available presets
+python3 -c "from reflex.embodiments import list_presets; print(list_presets())"
+
+# Use a custom config file instead
+reflex go --model smolvla-base --custom-embodiment-config ./myrobot.json
+```
+
+### `Model pull timeout`
+
+```bash
+# Use a mirror or set longer timeout
+export HF_HUB_DOWNLOAD_TIMEOUT=300
+reflex models pull smolvla-base
+```
+
+---
+
+## Quick Diagnostic Commands
+
+```bash
+# Full system check
+reflex doctor
+
+# Deploy-specific diagnostics
+reflex doctor --model ./reflex_export/ --embodiment franka
+
+# Drone diagnostics
+reflex doctor --model ./reflex_export/ --embodiment quadcopter
+
+# GPU info
+nvidia-smi
+python3 -c "import torch; print(f'CUDA: {torch.cuda.is_available()}, Device: {torch.cuda.get_device_name(0)}')"
+
+# ORT providers
+python3 -c "import onnxruntime; print(onnxruntime.get_available_providers())"
+```