PedroCuisinier2025/OBD2_panel_opel_2012

---
license: cc-by-4.0
---
# 📘 Dataset: OBD-II Telemetry – Opel Corsa 1.2 (2012)

Real-world automotive telemetry recorded from a 2012 Opel Corsa (A12XER, 84 hp), collected using an ELM327 OBD-II adapter and python-OBD.

---

## 📊 Overview

- **394,406 rows**
- **28 columns**
- Time-ordered samples from **2025-04-30 → 2025-12-02**
- Sampling frequency: **3–12 Hz** depending on PID latency
- Real OBD-II sensor readings + derived fields (fuel usage, torque, power, gear estimate)

Each row corresponds to a single OBD-II polling cycle during real driving:
city, suburban, and highway conditions.

---

## 📁 Columns

### 🔹 Core engine & time signals
- `timestamp`
- `RPM`
- `SPEED`

### 🔹 Throttle / load / air path
- `THROTTLE_POS`
- `RELATIVE_THROTTLE_POS`
- `ACCELERATOR_POS_D`
- `ENGINE_LOAD`
- `ABSOLUTE_LOAD`
- `INTAKE_PRESSURE`
- `INTAKE_TEMP`
- `MAF`

### 🔹 Fuel trims, O2 sensors, voltage, temperatures
- `SHORT_FUEL_TRIM_1`
- `LONG_FUEL_TRIM_1`
- `O2_B1S1`
- `O2_B1S2`
- `FUEL_LEVEL`
- `COOLANT_TEMP`
- `ELM_VOLTAGE`

### 🔹 Derived fuel fields
- `FUEL_USAGE_ML_MIN`
- `FUEL_USED_TOTAL_ML`
- `REAL_FUEL_USAGE_ML_MIN`
- `REAL_FUEL_USED_TOTAL_ML`

### 🔹 ML & helper fields
- `TORQUE`
- `POWER`
- `GEAR`
- `ENGINE_STATUS`
- `PREDICTED_FUEL_USAGE`
- `FUEL_USAGE_DIFF`
- `segment_id`
- `segment_file`

---

## ⛽ Fuel Usage Model

Fuel consumption is derived from the Mass Air Flow sensor (MAF):

```
fuel_g_s = MAF / 14.7
fuel_ml_s = fuel_g_s / 0.745
FUEL_USAGE_ML_MIN = fuel_ml_s * 60
```

- AFR = 14.7 (stoichiometric ratio)
- Fuel density = 0.745 g/ml
- Trapezoidal integration used for cumulative totals

---

## ⚙️ Gear Estimation Heuristic

OBD-II does not provide gear position.
This dataset includes a simple inference method based on `RPM / SPEED`.

```
if speed < 2 or rpm < 600:
return "N"
ratio = rpm / speed

if 130 >= ratio > 90: return "1"
if 90 >= ratio > 60: return "2"
if 60 >= ratio > 45: return "3"
if 45 >= ratio > 35: return "4"
if 35 >= ratio > 25: return "5"
return "?"
```

Limitations:

- fails during shifting, clutch press, rapid acceleration
- thresholds tuned for this specific vehicle
- included to allow ML gear classification research

---

## 🧪 Potential Machine Learning Tasks

### 🔧 1. Fuel consumption regression
Predict:
- `FUEL_USAGE_ML_MIN`
- `REAL_FUEL_USAGE_ML_MIN`

### ⚙️ 2. Gear classification (ML vs heuristic)
Using RPM, SPEED, load, throttle, trims, etc.

### 🚗 3. Driving style clustering
Eco / normal / dynamic driver profiles.

### 🛠 4. Anomaly detection
Identify irregularities in:
- O2 signals
- fuel trims
- MAF
- voltage
- coolant temperature

### 🔍 5. Time-series forecasting
Predict RPM, intake pressure, load, trims, etc.

---

## 🛠 Data Collection

- Car: **2012 Opel Corsa A12XER (84 hp)**
- Interface: **ELM327 adapter**
- Library: **python-OBD**
- Fast PIDs (~1s): RPM, SPEED, throttle position, load, MAF, trims, intake pressure/temp
- Medium (~15s): O2 sensors
- Slow (~30s): fuel level, coolant temp, system voltage
- Missing values are expected due to ECU latency and unsupported PIDs

---

## 📥 Load the dataset in Python

```python
from datasets import load_dataset

ds = load_dataset("PedroCuisinier2025/OBD2_panel_opel_2012")
df = ds["train"].to_pandas()
```

---

## 📜 License
This dataset is released under **CC-BY-4.0**.
You may use it freely, provided attribution is given.

---