PANORover: Autonomous Driving System Development Platform
summary
A small self-driving PANORover platform combines an Arduino, a Xilinx SoC with FPGA, a Mentor board, CAN and Ethernet links and a hardware-accelerated CNN that successfully brakes and avoids pedestrians in demos. The core claim is an extended Component Fault Tree (CFT) method that models both ISO 26262 failures and ISO 21448 SOTIF functional insufficiencies in one integrated model, implemented as a Capella/Arcadia MBSE view; analyses run minimal-cut-set checks, models are crafted manually today though partial automation is possible.
AI-written summary · about
chapters (9)
- 0:00Automotive industry growth
- 0:42PANORover platform
- 1:45Self-driving demo
- 3:30Hardware and AI overview
- 5:39AI safety assurance
- 7:23Fault tree analysis extension
- 9:35MBSE safety integration
- 12:47Hazard mitigation cuts
- 13:08Q&A
related talks
39mSoftware development challenges in automotive projects
03mThe DIY Frontier of Autonomous Driving
47mEmbedded Systems Architecture
46mPASTA: Portable Automotive Security Testbed with Adaptability
26mSelf Driving cars with Opencv
28mOpenAdx "Autonomous Driving Simulation" Testbed – hardware in the loop
20mSOAFEE-Based Software Defined Vehicle Development for Autonomous Driving
31mAutonomous driving, how Python drives Formula Racing cars