Upload README.md

README.md

@@ -6,16 +6,26 @@ tags:
   - numpy
   - obstacle-avoidance
   - simulation
+  - edge
+  - embedded
+  - industrial
+  - sensor-data
 pretty_name: Sim Driving MLP (NumPy)
 ---
 
-# Sim Driving MLP — 279-parameter obstacle-avoidance policy (NumPy)
+# Sim Driving MLP: a 270-parameter obstacle-avoidance policy (NumPy)
 
-A deliberately tiny neural network: **4 ultrasonic distances in, one driving command
-out.** 279 parameters, float32, trained and verified entirely inside our robot
-simulation studio. Small enough to read, small enough to run on an MCU-class device.
+**270 learnable parameters · 1.05 KB of weights · 0.0103 ms per inference · 1.95 KB peak inference memory · numpy only, no framework.**
 
-**Trained on simulation data from a virtual map — no real-world or customer data.**
+Inference numbers measured on a 12th-gen i7 laptop CPU, average over 100,000 runs.
+
+(An earlier revision of this card said 279 parameters. That count included the file's normalization and metadata arrays. The learnable weight and bias count is 270.)
+
+A deliberately tiny neural network: **4 ultrasonic distances in, one driving command out.**
+
+Small enough to read, small enough to run on an MCU-class device.
+
+**Trained on simulation data from a virtual map. No real-world or customer data.**
 
 ## Architecture
 
@@ -24,7 +34,7 @@ simulation studio. Small enough to read, small enough to run on an MCU-class dev
 | part | meaning |
 |------|---------|
 | input (4) | ultrasonic distances: **F**ront, **L**eft, **R**ight, **B**ack (normalized with the included `norm_mean` / `norm_std`) |
-| output (5 + 1) | 5 command logits — `FWD / LEFT / RIGHT / STOP / BACK` — plus 1 turning-angle regression head |
+| output (5 + 1) | 5 command logits (`FWD / LEFT / RIGHT / STOP / BACK`) plus 1 turning-angle regression head |
 
 ## Metrics (validation, virtual map)
 
@@ -33,8 +43,8 @@ simulation studio. Small enough to read, small enough to run on an MCU-class dev
 | command accuracy | **97.2 %** (best epoch 529 / 572) |
 | turning-angle MAE | ~6.9° |
 
-The `.npz` also embeds `meta_json`: the full 572-epoch training history
-(per-class accuracy, loss, angle MAE per epoch), so the training curve is inspectable.
+The `.npz` also embeds `meta_json`: the full 572-epoch training history (per-class accuracy, loss, angle MAE per epoch), so the training curve is inspectable.
+
 The plot below is drawn directly from that embedded history:
 
 ![Training history](assets/02_training_curve.png)
@@ -59,27 +69,25 @@ print(cmd, angle)
 
 ## The simulator and the target robot
 
-The studio stage where the model's four inputs are defined: ultrasonic sensors
-F/L/R/B with datasheet-based noise models, checked by the built-in self QA/QC
-checklist. On the right, the local LLM explains a warning from that checklist,
-citing the stage report as its basis. This screen is a frame from our
-[demo video (22 min, Korean)](https://youtu.be/ftsw_vbfw6E).
+The studio stage where the model's four inputs are defined: ultrasonic sensors F/L/R/B with datasheet-based noise models, checked by the built-in self QA/QC checklist.
+
+On the right, the local LLM explains a warning from that checklist, citing the stage report as its basis.
+
+This screen is a frame from our [demo video (22 min, Korean)](https://youtu.be/ftsw_vbfw6E).
 
 ![Sensor definition stage in the simulation studio](assets/05_sim_studio_sensors.png)
 
-A driving run on the 8 m × 8 m virtual map used for data collection
-(green: ultrasonic rays from the robot):
+A driving run on the 8 m × 8 m virtual map used for data collection (green: ultrasonic rays from the robot):
 
 ![Driving run on the virtual map](assets/06_virtual_map_run.png)
 
-And this exact model running in the studio's 3D evaluation stage. The left panel
-shows the live inference at the current step: the four sensor inputs (F/L/R/B, mm),
-the softmax over the five commands with FORWARD selected, and the angle head:
+And this exact model running in the studio's 3D evaluation stage.
+
+The left panel shows the live inference at the current step: the four sensor inputs (F/L/R/B, mm), the softmax over the five commands with FORWARD selected, and the angle head:
 
 ![The model driving in the 3D evaluation stage, with live inference panel](assets/07_nn_eval_3d.png)
 
-The target hardware: a tracked test robot with the ultrasonic sensors mounted on
-the hand, the same F-channel placement the simulator reproduces.
+The target hardware: a tracked test robot with the ultrasonic sensors mounted on the hand, the same F-channel placement the simulator reproduces.
 
 ![Tracked test robot with hand-mounted ultrasonic sensors](assets/04_real_robot.jpg)
 
@@ -87,25 +95,29 @@ the hand, the same F-channel placement the simulator reproduces.
 
 ![NCDTech 4-layer stack](assets/03_four_layer_stack.png)
 
-*(Diagram is in Korean; it is the same figure used on our website, demo video, and
-companion dataset.)*
+*(Diagram is in Korean; it is the same figure used on our website, demo video, and companion dataset.)*
 
-This model is a **layer-2 artifact** of our stack — an edge neural network verified
-through the 8-stage physics simulation workflow of our robot simulation studio.
-Every stage of that workflow runs self QA/QC (layer 3) and reports through an
-on-premise conversational LLM (layer 4); the record formats those layers produce
-are shown in our companion dataset:
+This model is a **layer-2 artifact** of our stack: an edge neural network verified through the 8-stage physics simulation workflow of our robot simulation studio.
+
+Every stage of that workflow runs self QA/QC (layer 3) and reports through an on-premise conversational LLM (layer 4).
+
+The record formats those layers produce are shown in our companion dataset:
 [NCDTech/human-gated-qaqc-knowledge-example](https://huggingface.co/datasets/NCDTech/human-gated-qaqc-knowledge-example)
 
 ## 한국어
 
-**초음파 4방향 거리(전·좌·우·후)를 넣으면 주행 명령이 나오는 279 파라미터짜리 작은
-신경망**입니다. 저희 로봇 시뮬레이션 스튜디오의 8단계 워크플로(데이터 수집 → 학습 →
-평가 → 물리 시뮬레이션 검증)를 통과한 계층 2(엣지 신경망) 산출물이며, **가상 맵
-시뮬레이션 데이터로만 학습**했습니다 — 실데이터·고객 데이터 없음.
+**초음파 4방향 거리(전·좌·우·후)를 넣으면 주행 명령이 나오는 270 파라미터짜리 작은 신경망**입니다.
+
+추론 1회 0.0103 ms, 추론 메모리 1.95 KB, 의존성은 numpy 하나 (12세대 i7 노트북 CPU에서 10만 회 평균 실측).
+
+(이전 판의 "279 파라미터"는 정규화·메타 배열까지 센 수치였고, 학습 가중치 기준 정확한 수는 270입니다.)
+
+저희 로봇 시뮬레이션 스튜디오의 8단계 워크플로(데이터 수집 → 학습 → 평가 → 물리 시뮬레이션 검증)를 통과한 계층 2(엣지 신경망) 산출물이며, **가상 맵 시뮬레이션 데이터로만 학습**했습니다 (실데이터·고객 데이터 없음).
 
 출력은 주행 명령 5클래스(전진/좌회전/우회전/정지/후진) + 회전각 회귀 1개.
-검증 정확도 97.2 %, 각도 오차 약 6.9°. 파일 안에 정규화 통계와 572 에포크 학습
-이력 전체가 함께 들어 있어 학습 곡선을 그대로 확인할 수 있습니다.
+
+검증 정확도 97.2 %, 각도 오차 약 6.9°.
+
+파일 안에 정규화 통계와 572 에포크 학습 이력 전체가 함께 들어 있어 학습 곡선을 그대로 확인할 수 있습니다.
 
 Learn more: https://huggingface.co/NCDTech · https://www.ncdtech.org · Demo video: https://youtu.be/ftsw_vbfw6E