Gogada Prameela · B.Tech ECE · Class of 2027Software & Embedded.

01 · Final-year project · Sep 2026 – Apr 2027 · in progress

Automated Validation and Anomaly Detection for In-Vehicle CAN Networks

A bench-top car network: a Raspberry Pi 4 gateway and two ESP32-S3 boards acting as engine and ABS control units talk over a real CAN bus. A C program on Linux SocketCAN decodes the signals from a DBC file, and an automated pytest suite injects faults and produces PASS/FAIL reports.

• Three-node CAN network on the bench: Raspberry Pi 4 gateway plus two ESP32-S3 ECUs (engine and ABS).
• C program on Linux SocketCAN that decodes speed, RPM and temperature from a DBC file and flags out-of-range values.
• Automated pytest suite with fault injection (invalid values, missing messages) that generates PASS/FAIL reports.
• Next: machine-learning detection of spoofed CAN messages, and decoding a 433 MHz tyre-pressure sensor with an RTL-SDR.

1. Phase 1

   CAN + validation

   In progress

2. Phase 2

   ML anomaly detection

   Next

3. Phase 3

   Radio + final report

   Later