Upload 5 files

.gitattributes

@@ -33,3 +33,5 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+assets/02_real_robot.jpg filter=lfs diff=lfs merge=lfs -text
+assets/03_robot_in_action.png filter=lfs diff=lfs merge=lfs -text

README.md

@@ -1,3 +1,106 @@
 ---
 license: cc-by-4.0
+tags:
+  - robotics
+  - mlp
+  - numpy
+  - obstacle-avoidance
+  - real-robot
+pretty_name: Real-Robot Driving MLP (NumPy)
 ---
+
+# Real-Robot Driving MLP — the 116-parameter policy that drove our robot on video
+
+This is the smaller, older sibling of
+[NCDTech/sim-driving-mlp-numpy](https://huggingface.co/NCDTech/sim-driving-mlp-numpy):
+a 116-parameter MLP trained not in simulation but on **real driving sessions** of our
+tracked test robot, collected on the test floor through our desktop studio.
+
+**This exact file is the one loaded and driving in the video below.** We traced it
+frame by frame: the load dialog in the recording shows this very filename.
+
+▶ **[Watch it drive (3:18, Korean)](https://youtu.be/6DJX0T6qrtM)** — synthetic data
+check, analysis, training, weight transfer to the robot, and neural-network driving,
+end to end.
+
+## Architecture
+
+![Architecture: 6 → 8 → 5](assets/01_architecture.png)
+
+| part | meaning |
+|------|---------|
+| input (6) | ultrasonic distances **F**ront / **L**eft / **R**ight, plus their frame-to-frame **deltas** (ΔF, ΔL, ΔR) — distance tells "how far", delta tells "getting closer or not" |
+| hidden (8) | one dense layer, leaky ReLU (α = 0.01) |
+| output (5) | command logits: `FWD / LEFT / RIGHT / STOP / BACK` |
+
+Normalization statistics (`norm_center`, `norm_scale`) are included in the file.
+
+## Metrics — an honest early scorecard
+
+| command | accuracy |
+|---------|----------|
+| FORWARD | 76 % |
+| LEFT | 85 % |
+| RIGHT | 83 % |
+| STOP / BACK | **0 %** |
+| overall | 49 % |
+
+We publish these numbers as they are. The driving commands (F/L/R) worked well
+enough to drive the robot in the video; STOP and BACK scored zero because the
+training sessions barely contained those actions (class imbalance in a few minutes
+of real driving data). Fixing exactly this kind of gap is why our current systems
+run self QA/QC on every stage — the class-distribution check that would have
+flagged this dataset is now built in.
+
+## Load and run (NumPy only)
+
+```python
+import numpy as np
+
+d = np.load("real_robot_mlp.npz")
+x = np.array([[300.0, 150.0, 400.0, -5.0, -20.0, 0.0]])  # F, L, R, dF, dL, dR
+x = (x - d["norm_center"]) / d["norm_scale"]
+
+h = x @ d["W0"] + d["b0"]
+h = np.where(h > 0, h, 0.01 * h)          # leaky ReLU
+y = h @ d["W1"] + d["b1"]
+
+print(["FWD", "LEFT", "RIGHT", "STOP", "BACK"][int(np.argmax(y))])
+```
+
+The whole file is 2.4 KB. It runs on anything that runs NumPy, and the original
+target was an MCU-class robot controller.
+
+## The robot it drove
+
+![Tracked test robot with hand-mounted ultrasonic sensors](assets/02_real_robot.jpg)
+
+![The robot driving on the test floor, monitored live by our control dashboard](assets/03_robot_in_action.png)
+
+## Sim and real, side by side
+
+| | this model | [sim-driving-mlp-numpy](https://huggingface.co/NCDTech/sim-driving-mlp-numpy) |
+|---|---|---|
+| trained on | real driving sessions (test floor) | virtual-map simulation |
+| inputs | 3 distances + 3 deltas | 4 distances (F/L/R/B) |
+| size | 6→8→5, 116 params | 4→16→8→6, 279 params |
+| extra head | none | turning-angle regression |
+| date | 2026-04 | 2026-05 |
+
+Together they show the path we actually took: drive the real robot first, feel the
+data problems, then build the simulation studio with self QA/QC wired into every
+stage.
+
+## 한국어
+
+**영상 속에서 실제로 로봇을 몰던 바로 그 가중치 파일**입니다 (영상의 파일 열기
+장면에서 파일명을 추적해 확인). 시뮬 모델의 형뻘로, 가상 맵이 아니라 **실제 주행
+데이터**로 학습했습니다: 초음파 3방 거리 + 변화량 3개를 넣으면 주행 명령 5종이
+나오는 116 파라미터 신경망(2.4 KB)입니다.
+
+성적표는 있는 그대로 공개합니다: 주행 명령(전진 76%·좌 85%·우 83%)은 영상처럼
+실주행이 됐지만, **정지/후진은 0%** — 몇 분짜리 실주행 데이터에 그 동작이 거의
+없었기 때문입니다(클래스 불균형). 바로 이런 구멍을 잡으려고 지금의 저희 시스템은
+모든 단계에 셀프 QA/QC(클래스 분포 검사 포함)를 심었습니다.
+
+Learn more: https://huggingface.co/NCDTech · https://www.ncdtech.org · 주행 영상: https://youtu.be/6DJX0T6qrtM

real_robot_mlp.npz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:697a394419f2cbb4e98de4edfa1c2c902efa94c4a3aa4cf7e4413a57807c5305
+size 2448