Firmware.RE: Firmware Unpacking, Analysis and Vulnerability-Discovery as a Service
summary
A large-scale pipeline unpacks and recursively carves firmware, yielding 32K firmware images expanded into 1.7M files, with automated static checks (fuzzy-hash clustering, modified binwalk-style unpackers, password-hash cracking) and cloud worker nodes. Correlating binaries and SSL keys with ZMap SSL collection exposes shared vulnerable components; the dataset links 38 new CVEs across 693+ firmware images and over 140K Internet-accessible devices. A public web upload/analysis interface and extracted-file reports are operational, though full dynamic confirmation of each finding remains manual.
AI-written summary · about
chapters (10)
- 0:00Introduction and Motivation
- 11:37Challenges in Analysis
- 14:48Unpacking and File Carving
- 22:25The Analysis Platform
- 25:31Data Set Overview
- 27:29Firmware Correlation Analysis
- 30:44Findings and Results
- 33:44Demonstration and Case Study
- 39:15Website Features and Future
- 44:29Conclusion
related talks
36m Hashing the $#!+ out of firmware
01h 08mAutomated Dynamic Fireware Analysis At Scale: A Case Study on Embedded Web Interfaces
33mIdentifying Multi-Binary Vulnerabilities in Embedded Firmware at Scale
39mNight of the living vulnerabilities
37mThe Firmware Supply-Chain Security Is Broken: Can We Fix It?
29mFragile by Design: Large-Scale Evidence of Supply Chain Risk
45mOutside the box: pwning IoT devices through their applications
44mCome Join the CAFSA - Continuous Automated Firmware Security Analysis