projects | Emad Mahmodi

work

Negaremap

A professional and specialized Iranian software for ECU calibration (remapping), intended for lawful and emissions-compliant use.

PISAD ECU Full Simulator

A high-fidelity PowerPC (PPC) emulator for full ECU firmware analysis, defensive automotive security research, and assessment of SecurityAccess (SID 0x27) mechanisms.

ECUProg

Senior ECU Programmer & Security Expert at Negar-Khodro| Developer of ECUProg v2 – Advanced Automotive Programming Tool

Security Assessment of Software Protection in Diagnostic Tools

Assessing protection layers in automotive diagnostic software using x32dbg and anti-tamper analysis.

Security Assessment of Diagnostic Scanner Ecosystems

Defensive assessment of the security posture of Android-based OBD2 scanners and their Bluetooth VCI communications.

Automotive Protocol Analyzer & Monitoring Tool

A high-performance monitoring app built with Qt/QML and Python to analyze Saleae Logic exports for UDS and OBD-II diagnostics.

Advanced Reverse Engineering & Hardware Interfacing of a Multi-Core Automotive ECU

A defensive study of multi-core firmware architecture, SecurityAccess (UDS 0x27) robustness, and live CAN-Bus monitoring for ECU diagnostics.

Reverse Engineering of an Automotive Engine Control Unit

Defensive full-stack analysis of an engine ECU, from hardware interfacing to line-by-line binary analysis and logic understanding.

LLM-Assisted Binary Code Understanding & Reverse Engineering

Benchmarking and developing RAG pipelines for automated binary analysis and assembly-to-C translation using LLMs.

Binary-Analysis LLM Copilot — Studying ReCopilot & a Lightweight RAG Baseline

A study of ReCopilot (Chen et al., 2025) — an expert LLM for binary analysis — together with my own independent, lightweight retrieval-augmented (RAG) reimplementation for function-name recovery.

GPU-Accelerated Graph Analytics for Irregular Security Graphs

A reproducible empirical study of GPU acceleration strategies (BFS, PageRank, WCC, Louvain) on skewed, power-law graphs — five optimization tiers from sequential C++ to load-balanced CUDA, benchmarked from 8M to 268M edges, with every speedup attributed to one named technique and every negative result kept.

Cross-Domain Generalization of Anomaly Detection on Network-Intrusion Data

A PyTorch reimplementation of my drift-aware, uncertainty-weighted online ensemble (ESWA 2020), evaluated under real distribution shift — trained on one attack distribution, tested on another.

RL-Guided Fuzzing for Binary Vulnerability Detection — an Independent Reimplementation and 13-Agent Study of VSGFuzz

An end-to-end independent reimplementation of VSGFuzz (Array, 2025) — graph-embedding vulnerability prediction, block-value seed scoring, and reinforcement-learned mutation scheduling — together with the controlled multi-seed benchmark the paper never ran: thirteen RL agents on one MDP, one reward, the same seeds.