Safety-Critical Systems from the Inside - NDC Oslo 2020

Explore safety-critical systems in this NDC Oslo 2020 conference talk. Delve into techniques for ensuring safety in systems where malfunctions can lead to dangerous accidents. Learn how laws of physics are applied to achieve fail-safety, the benefits of using multiple CPUs for defect detection, and the importance of separating less critical tasks. Discover methods for ensuring communication safety, creating useful and up-to-date documentation, and how programming language choice impacts safety. Investigate the possibility of formally proving code free of runtime errors. Gain insights into safety integrity levels, redundancy, mixed criticality, data corruption prevention, formal proofs, and effective documentation practices. Understand the roles of people and processes in safety-critical projects, root cause analysis, and feedback loops. Access additional resources to further explore engineering safer systems.

Intro
MACIEJ GAJDZICA
SYSTEM WHOSE MALFUNCTI CAN LEAD TO
SAFETY VS SECURITY
SAFETY VS RELIABILITY
Risk: cutting fingers
Risk: burning everything
SAFETY INTEGRITY LEVEL SIL
FAIL SAFE
CPU ERRORS
HOW TO DETECT CLOCK FAIL
REDUNDANCY
SAFE COMMUNICATION
PROBLEM
SOLUTION
MIXED CRITICALITY
DATA CORRUPTION
uint32_t const value; uint32_t const value_inv
FORMAL PROOF
ADA SPARK
LANGUAGE SUBSETS
EFFECTIVE DOCUMENTATION
VERSION MANAGEMENT
PEOPLE AND PROCESSES
ROOT CAUSE ANALYSIS
FEEDBACK LOOP
Engineering a Safer World
PROJECT ROLES - SIL4
ADDITIONAL RESOURCES