Crash Test Signal
Industrial AI / MLOps
An AWS Lambda inference service for validating crash-test sensor channels with real-time signal processing.
Overview
Crash-test sensor validation needs a deployable inference path that combines signal imagery and structured channel features.
The problem
The model and preprocessing path must run consistently inside an event-driven cloud service.
My contribution
Implemented a containerized Lambda inference flow with S3 integration, Pydantic event schemas, signal processing, and a hybrid PyTorch model.
Technical highlights
- 01ResNet18 hybrid model combining signal images with tabular features
- 02Containerized Lambda deployment
- 03Pydantic schemas for type-safe event handling