adipu/Guido.Tech
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

update dreadmed

Browse Source
This commit is contained in:
rishubm
2025-10-19 11:25:30 -05:00
parent 154283e7c6
commit 37aaef3f62
3 changed files with 41 additions and 90 deletions
Download Patch File Download Diff File Expand all files Collapse all files

117
README.md
View File

@@ -1,94 +1,54 @@
# HPCSimSite # Guido.tech:
HPC simulation site ## An F1 AI Race Engineer System
# F1 Virtual Race Engineer — Enrichment Module Real-time F1 race strategy system combining telemetry enrichment with AI-powered strategy generation. The system receives lap-by-lap telemetry from Raspberry Pi hardware, enriches it with performance analytics, and generates dynamic race strategies using Google Gemini AI.
This repo contains a minimal, dependency-free Python module to enrich Raspberry Pi telemetry (derived from FastF1) with HPC-style analytics features. It simulates the first LLM stage (data enrichment) using deterministic heuristics so you can run the pipeline locally and in CI without external services. ## Architecture
## What it does The system consists of two main services:
- Accepts lap-level telemetry JSON records.
- Produces an enriched record with:
- aero_efficiency (0..1)
- tire_degradation_index (0..1, higher=worse)
- ers_charge (0..1)
- fuel_optimization_score (0..1)
- driver_consistency (0..1)
- weather_impact (low|medium|high)
## Expected input schema 1. **Enrichment Service** (`hpcsim/`) - Port 8000
Fields are extensible; these cover the initial POC. - Receives raw telemetry from Raspberry Pi
- Enriches data with tire degradation, pace trends, pit window predictions
- Forwards to AI Intelligence Layer via webhook
Required (or sensible defaults applied): 2. **AI Intelligence Layer** (`ai_intelligence_layer/`) - Port 9000
- lap: int - WebSocket server for real-time Pi communication
- speed: float (km/h) - Generates race strategies using Google Gemini AI
- throttle: float (0..1) - Sends control commands (brake bias, differential slip) back to Pi
- brake: float (0..1) - Web dashboard for monitoring
- tire_compound: string (soft|medium|hard|inter|wet)
- fuel_level: float (0..1)
Optional: ## Quick Start
- ers: float (0..1)
- track_temp: float (Celsius)
- rain_probability: float (0..1)
Example telemetry line (JSONL): ### Prerequisites
{"lap":27,"speed":282,"throttle":0.91,"brake":0.05,"tire_compound":"medium","fuel_level":0.47}
## Output schema (enriched) - Python 3.9+
Example: - Google Gemini API key
{"lap":27,"aero_efficiency":0.83,"tire_degradation_index":0.65,"ers_charge":0.72,"fuel_optimization_score":0.91,"driver_consistency":0.89,"weather_impact":"medium"}
## Quick start ### 1. Install Dependencies
### Run the CLI
The CLI reads JSON Lines (one JSON object per line) from stdin or a file and writes enriched JSON lines to stdout or a file.
```bash ```bash
python3 scripts/enrich_telemetry.py -i telemetry.jsonl -o enriched.jsonl # Install enrichment service dependencies
pip install -r requirements.txt
# Install AI layer dependencies
pip install -r ai_intelligence_layer/requirements.txt
``` ```
Or stream: ### 2. Configure Environment
Create `ai_intelligence_layer/.env`:
```bash ```bash
cat telemetry.jsonl | python3 scripts/enrich_telemetry.py > enriched.jsonl GEMINI_API_KEY=your_gemini_api_key_here
ENVIRONMENT=development
FAST_MODE=true
STRATEGY_COUNT=3
``` ```
### Library usage ### 3. Run the System
```python **Option A: Run both services together (recommended)**
from hpcsim.enrichment import Enricher
enricher = Enricher()
out = enricher.enrich({
"lap": 1,
"speed": 250,
"throttle": 0.8,
"brake": 0.1,
"tire_compound": "medium",
"fuel_level": 0.6,
})
print(out)
```
## Notes
- The enrichment maintains state across laps (e.g., cumulative tire wear, consistency from last up to 5 laps). If you restart the process mid-race, these will reset; you can re-feed prior laps to restore state.
- If your FastF1-derived telemetry has a different shape, share a sample and we can add adapters.
## Tests
Run minimal tests:
```bash
python3 -m unittest tests/test_enrichment.py -v
```
## API reference (Enrichment Service)
Base URL (local): http://localhost:8000
Interactive docs: http://localhost:8000/docs (Swagger) and http://localhost:8000/redoc
### Run the API server
```bash ```bash
python3 scripts/serve.py python3 scripts/serve.py
@@ -122,16 +82,7 @@ Accepts raw Raspberry Pi or FastF1-style telemetry, normalizes field names, enri
Example request: Example request:
```bash ```bash
curl -s -X POST http://localhost:8000/ingest/telemetry \
-H "Content-Type: application/json" \
-d '{
"LapNumber": 27,
"Speed": 282,
"Throttle": 0.91,
"Brakes": 0.05,
"TyreCompound": "medium",
"FuelRel": 0.47
}'
``` ```
Response 200 (application/json): Response 200 (application/json):

2
ai_intelligence_layer/README.md
View File

@@ -1,4 +1,4 @@
# F1 AI Intelligence Layer r# F1 AI Intelligence Layer
**The core innovation of our HPC-powered race strategy system** **The core innovation of our HPC-powered race strategy system**

12
scripts/simulate_pi_websocket.py
View File

@@ -9,7 +9,7 @@ Connects to AI Intelligence Layer via WebSocket and:
4. Generates voice announcements for strategy updates 4. Generates voice announcements for strategy updates
Usage: Usage:
python simulate_pi_websocket.py --interval 5 --ws-url ws://localhost:9000/ws/pi --enable-voice python simulate_pi_websocket.py --interval 5 --ws-url ws://192.168.137.134:9000/ws/pi --enable-voice
""" """
from __future__ import annotations from __future__ import annotations
@@ -306,7 +306,7 @@ class VoiceAnnouncer:
class PiSimulator: class PiSimulator:
"""WebSocket-based Pi simulator with control feedback and voice announcements.""" """WebSocket-based Pi simulator with control feedback and voice announcements."""
def __init__(self, csv_path: Path, ws_url: str, interval: float = 60.0, enrichment_url: str = "http://localhost:8000", voice_enabled: bool = False): def __init__(self, csv_path: Path, ws_url: str, interval: float = 60.0, enrichment_url: str = "http://192.168.137.134:8000", voice_enabled: bool = False):
self.csv_path = csv_path self.csv_path = csv_path
self.ws_url = ws_url self.ws_url = ws_url
self.enrichment_url = enrichment_url self.enrichment_url = enrichment_url
@@ -719,14 +719,14 @@ async def main():
parser.add_argument( parser.add_argument(
"--ws-url", "--ws-url",
type=str, type=str,
default="ws://localhost:9000/ws/pi", default="ws://192.168.137.134:9000/ws/pi",
help="WebSocket URL for AI layer (default: ws://localhost:9000/ws/pi)" help="WebSocket URL for AI layer (default: ws://192.168.137.134:9000/ws/pi)"
) )
parser.add_argument( parser.add_argument(
"--enrichment-url", "--enrichment-url",
type=str, type=str,
default="http://localhost:8000", default="http://192.168.137.134:8000",
help="Enrichment service URL (default: http://localhost:8000)" help="Enrichment service URL (default: http://192.168.137.134:8000)"
) )
parser.add_argument( parser.add_argument(
"--csv", "--csv",